nthe : ns eps i Pe WS Wa vite i bbb ea iti aehe 3 i atin > ake Ch aha Leg EVRA PEA, ; PSD fa HRC Saat Dk BE reece \ AR aA r EIN EEC RL ILO TOE o ena Cade ay aL ge Vb ats iN ih Bae wand Hae err RO RET Er COU veil tidy ye Awa Ae Tv ai ee e y Dee Mey sae cee f aN 4 ; somal Pan mel a 3 VN VAN BN cated deh sage i tte t ‘ ( ay ekay at ar Aa x i HEANOR ARVIN REA het : ae RETNA Shere the CE on a G a Le Se a eye ib : aan BINA, Feed ANIA PSE TERE NERDS Ne 2. sevege reared fy Cash See E nase HeAVAN py ges Pen AN HP TENS EGY FEAL dS et ~~ wan sm Soba Oe ee fy TARY Se aie fas + Ob yey wietab eed eae mi a eee Siti Nt enen RAGA Seen yy (RATAN . i ye eh, N Le Warr iv ape das yy Th eae y Oe si TY Wiha SN eR tai BEN EU loth Paves OPENER CNET Lee LEM MSIN he DEM EE Ade POEM AT EME WARS ip wy pire AN te) DE Dep yeig be \ a ; ae hae ete, aego wel dyin Joa eae Eby tae AVI Jd Peery RUN Er DY SEY Yordibe Ni) BOK AEN) Nr AM ied Pee oye eu Ris vy ray 27 nway WSS Arr ak yiegWey 5 dha” Bye dy Es ai Lt ek ay sheteaan ven saan ean Ns RMS ha Rohe toon SPEER Rei rT ent a (rb yang” eMbmte jek of Fy wt a aH Nin aae a ig bad ev eerste star pit 1 Rilnwb ay i meth a B ay de Rng titi why ’ Fe EuN INTE HDTV AV ch gcd byes ban yoy ah LIN fh Said ’ eek aya ; at tei ; s Teed tor bed YA Lv 1 A ON YY ay YSBh TEN die 48 7 Wl Seeetth mayen el, Dn Dhaest a wan ok we ANE wrod eweteN hy yur yeate A UNA ae A tn Wa tk A 2B 1 yon he sebedfh latin gb Uliea 8 hed ints pide ; els: we hed eis) ai Bs Hie tha Ee hay See eP eS Ke Me PRA aT govt TMF FORE TGS ANG TS WAN ie ARS RE dee Mi fh» HN He te CMEC Het : D d ae ea htt ty th seat GD of Ay Dhan ab a " i % i 4 tests nel sain NT nL a ” tne tawny startet I day W.uy) af Y DW HTN EG at a ny Na ‘e 2 a ebyek walk cb y: par athe sit Ut Bans ME See bas eh REE BY 1 $ hha Ba nt voy MEDIA vine! PH ey , q Sa MUU ENTE a Vere SA PR SS eT ee a Vid ev ee gba eke Sea a Bieri r Mi panes an NE ast med ete a oxen pone barber te +e tively ‘ rere é oat 5 vetaPetaitesily De ee et cee ee eta : ‘ cS Hee OD RE ee Sa hes if OS en y ‘ i i Pas res youn ket Ee MCN er ren or ek Let wee a ; wee CRESS tobe WY TAL NS = q , ¢ BAN eta ler euiey bal ‘ Heat yw ATA ae Sew hon ee nab Sie VIER AVY NN eS Gb bl wares iy CIS dsb vp et ne Yop MY asta YR STD AE EW Here wr dB ND C0 vib aie yf Solera cba aS Hu LCs So 8 gt doy sans nd BIN OY , SHAT ba 8 > Ni : ; 5 Vs we ATE Peal anavin pbx eee wove saci we eee te tes waned ve oh Gages EDS SV Ce WORE NE ANS EN ME TNT UU BON YRS Ut ey Mae kota whys yes Mat al os he de th Steed potty oe fe eth ity Cie dset ‘ Seley ¥ Lye he aes” ee ee yo Mee ter hen eR ODT C Ten RNR =e i Bh pape dh RBI EY Ms PE TOT de ey 4 é . bya nee HIDE PE Dap TIS Shy SAAN EE MEN giao Pe eS ERIE EMD eat eer “gh ee he > vee deve pgs er dae hue ny 4 . shotanh Petes Tee Cee ey ay . Saba ae ABty t tlds pd dh pA dhs b te (bh Ne \ phew ded hal ange sesryi - 4 A ; i aay Mug geen Nev fea an Eye vpn d ha Lbs es r ay Lh en veh Vento vase Maire ry WS de vig Soares ree setae hip IO eT hey a eee y Be yawe een ras Si wha we Vi rani SA ea gee aL dol ach HA ene dedi *AVIV EH an eerie Sark haem ewe Seo SYA me : HNN x FE ua the eais qa siyduakeg OLN8 erent wernt ar vars Sete rely eeu ; OR ya Aiba wea tab Say Td : ‘ ee ae hoes . . \ eee tee ane : fi % 9% vey rene ; Stuf WLAN PSS sae oe wate: WLU are t ep eee Wy MS RITE VT ATOM ve eNO i Y ee a peed Saree ea tee rien RR eee eg curere rete veh Gares eutat Te LAe CURRANT eink MODE Y TN yauareds ibe - SN GY Wee kige gs Ow : ¢ ; : E eee) ee “ h om ¢ oxy Lod Vreete SV de yokte dod 4 $ apa en ove Fe? : 7 " WSO PN ese aca Ser eran pe Oss, er he Pens ‘ Ret: Y Veneta te eegy eres He baty nba ae sar) wit a teatlediy yaa ae “ ada keewew aN, Sash oD CLEA Belen ee aay Ser State SE Ga tana S eee FICONRINT Fas te daha BS eater etre ten Ore Se BE caey Ceca ewan Arete as Sgpathoug spy Betere : whasbety oS nese ¢ fas au Stasi SAR x ne es Spie NG Sip P PAT SR Sake vee We ces _seadeat oh . ave » * =e Doran alate See = Ree : ’ UME eit one Waleed VES aking estiiientlnccctg li f SMe dot trv ’ Ove de ediby dave are irey, san tetsicitin-edartieid PAW ae ak he EU OvMR Cre ee A whee Mew ea 4 Se te ie on re ea ; STD Nseue pent uae z Mbt neato eS NEN Finds rhgead svt Svbge tad ek, Sy Inca ky een Spoiees 2 ewes Ly NAVAL Pen de teed SNe eh tree gd Day Oya a aks 998 " i} , was td Sein at vat fas gov sy aye 7 aepteckabee rt PRS Sa ae Gi eka rut sy ” 4 % u pth Shed agaty ote ws WhUMe Ya ey ee wo aR ate ts we ee ee Wve > srg deys : ee Sen ee rem dy ete Peet pean Fs 2p ea ESAS 9 sty! Jip tutias vox tein { reds gneas $ Nee tary 7 . > aad DOIN, eds Sepa iotreknite “Eat tiene 4 ‘ Pind cid othe vein eerie ; a } Sn ern . = Woks bal £ i wy a f LOAM cman ech wy aS yon Se aes aint dan ater gee eS ra Fe eas at F b grew wrt Nas 4 whet hatte RB Bote ae SEP eee n:dVaip 2 $i Haphawe aly a esa ae ¥ : * Ny d Sreh E are es A acne f. corere we ee 4 Lys ; ‘ aa) att . . . pepe rudy ne ttont ea ra al ’ i . F ‘ . Chase Fara te hae : . Brain ae a? eee BO Rw r | { ah wey het ee b eet eave’ eres thane et hese sap mer adign aud hag PRED Eee eh edd ha gktvewipe’ eyo ay a pcan 7 » ‘ . . “ : Hee Nnlau nea bantie = Loe eae a oevee es 4 Me ee eee ee teehee ‘ athe th on obetorth ea erg ee ee Hedin Hee gy ter rant 0% ” . oN ee teen ent iim ge va OLN AW bilan Si ire We a os WH Z WATSON! Journal of the Botanical Society of the British Isles MAR 0.9 1998 LIBRARIES Volume 22 Part 1 February 1998 Editors: M. Briggs, J.R. Edmondson, D.L. Kelly C.D. Preston, B.S. Rushton, M.N. Sanford Botanical Society of the British Isles Patron: Her Majesty Queen Elizabeth the Queen Mother Applications for membership should be addressed to the Hon. General Secretary, c/o Department of Botany, The Natural History Museum, Cromwell Road, London, SW7 5BD, from whom copies of the Society’s Prospectus may be obtained. Officers for 1997-98 President, Mr D. A. Pearman President-elect, Mrs M. Briggs Vice-Presidents, Mr R. G. Ellis, Dr C. D. Preston, Dr R. J. Gornall, Mr M. Walpole Honorary General Secretary, Mr R. G. Ellis Honorary Treasurer, Mr M. E. Braithwaite Editors of Watsonia Papers and Notes, J. R. Edmondson, D. L. Kelly, B. S. Rushton*, M. N. Sanford Plant Records, C. D. Preston Book Reviews, C. D. Preston Obituaries, M. Briggs “Receiving editor, to whom all MSS should be sent (see inside back cover). © 1998 Botanical Society of the British Isles The Society takes no responsibility for the views expressed by authors of Papers, Notes, Book Reviews or Obituaries. The cover illustration of Potentilla anglica Laich. (Trailing Tormentil) was drawn by Rosemary Wise. — Watsonia 22: 1-19 (1998) | The independent gametophytic stage of Trichomanes speciosum Willd. (Hymenophyllaceae), the Killarney Fern and its distribution in the British Isles F. J. RUMSEY, A. C. JERMY Department of Botany, the Natural History Museum, Cromwell Road, London, SW7 5BD and E. SHEFFIELD School of Biological Sciences, University of Manchester, Oxford Road, Manchester, M13 9PT ABSTRACT Uniquely amongst European ferns Trichomanes speciosum Willd., the Killarney Fern, has perennial, gemmiferous gametophytes which may grow and persist in the absence of the sporophyte generation. The presence of widespread independent colonies of the gametophyte generation and their habitats is briefly described and their distribution in the British Isles documented. The conservation implications of this unique situation are discussed. KEyYworbs: gametophyte, sporophyte, pteridophyte, Atlantic cryptogamic community, ecology, conservation biology. INTRODUCTION HISTORICAL Trichomanes speciosum Willd., the Killarney Fern or Bristle Fern has been described as “‘one of the rarest and most celebrated species in the British flora” (Ratcliffe et al. 1993). The species’ great rarity, beauty and image of tropical incongruity have acted to generate a potent mythology, only paralleled by that of some orchids. It is one of a large genus of filmy-ferns showing their greatest diversity in tropical montane rain forests. Trichomanes speciosum was originally described from plants collected in the Canary Islands and is restricted now to those and other islands of the Macaronesian archipelago and a few relict sites in Europe. In Britain considerable and understand- able secrecy has surrounded the location of this fern, following the wholesale depredation of localities to adorn the drawing rooms of the Victorian upper classes (Allen 1969). Trichomanes speciosum has a long history of records, a resumé of which is given below. The first ever collection of 7. speciosum in the British Isles was that by Dr Richard Richardson in 1724 (Ray 1724), by a spring head on Bell Bank, Bingley, Yorkshire, voucher material of which can be seen in the Sloane Herbarium at BM - (HS. 145, f. 9; H.S. 302, f. 66). Within 50 years its original site had been destroyed and T. speciosum was thought to have died out but was refound in the early 1780s (Bolton 1785; Teesdale 1800) and again brought into cultivation. Unfortunately this was probably responsible for the plant’s demise, as from about 1785 T. speciosum was not recorded as being seen again and was, once again, considered extinct. The botanical exploration of Ireland in the first years of the 19th century, however, revealed this species to be more widespread and achieving abundance - at some sites in the extreme south-west, most notably around the Killarney area of Co. Kerry (v.c. H1-2). This fact is commemorated in its current vernacular name, although to many Victorian botanists it was the “Irish Fern’’. Its potential for specialised culture, already developed for plants demanding similar high humidity brought in from New Zealand and elsewhere, was quickly seen, and plants were sold through professional outlets in Britain, and by rural entrepreneurs in Ireland. 2 F. J. RUMSEY, A. C. JERMY AND E. SHEFFIELD As the nineteenth century advanced, the advent of the Wardian case enabled the elite of London to have Killarney ferns in their drawing rooms. Almost from its discovery, Trichomanes speciosum was under threat in Britain and Ireland. Collection and subtle but crucial disturbance to microhabitats through visitation still arguably pose the greatest threats to this species’ survival (Wigginton, in press). The necessary caution surrounding the plant’s localities has, however, made it difficult to establish losses and gains, as the history of the species within its few sites, e.g. in terms of extent, performance and fertility is, where known, unpublished, or so vaguely identified that its worth is regrettably limited. Even past and present distribution in broader terms is difficult to establish accurately, posing further complications for those seeking to understand the environmental factors acting to limit the species’ distribution and explaining its dispersal. This is of increased importance given recent discoveries which have revealed a disparity in the distributional extent and amount of the two phases of the life cycle. THE GAMETOPHYTIC STAGE OF THE LIFE CYCLE Trichomanes speciosum is unique amongst European ferns in that its gametophytic generation, the sexual or gamete-bearing phase of the life cycle, is not only perennial but produces specialized structures for its vegetative propagation (gemmae), allowing the potential development of extensive stands of this usually overlooked generation. Originally described and illustrated from cultivated material over a century ago (Goebel 1888), the distinctive filamentous gametophyte of this species appears to have been completely overlooked in the field prior to the finds reviewed here. The fact that fern gametophytes can establish and reproduce themselves independently of the sporophyte has been known for some time, and the, now classic, examples from Eastern U.S.A. of gametophytes of Hymenophyllaceae, and other ferns in the families Grammitidaceae and Vittariaceae that have in some instances spread several hundreds of kilometres beyond the ranges of the sporophyte or persist in their complete absence, have been well-documented (Farrar 1967; Farrar et al. 1983). Farrar (1985) has shown that populations of the gametophyte of N. American Trichomanes boschianum Sturm, the United States endemic “equivalent” of 7. speciosum, exist up to 40 km distance from the nearest sporophytes. In addition, a further more widespread and numerous T7richomanes taxon, found in the eastern United States and initially thought to be the gametophyte of 7. petersii A. Gray, has been shown (by comparison of enzymes) to be a new species, 7. intricatum Farrar, currently known only as its gametophyte generation (Farrar 1992). Preliminary investigation elsewhere suggested independent 7richomanes gametophyte populations may be a widespread phenomenon in temperate areas (Rumsey & Sheffield 1996). During a sabbatical visit to the U.K. late in 1989, Farrar discovered gametophytes he recognized to be those of a Trichomanes species at two sites in the English Lake District (Rumsey ef al. 1990). These were compared isozymically with a range of Trichomanes species and gave banding patterns identical to, and characteristic of, 7. speciosum (Rumsey et al. 1993; Rumsey-1994). As records increased of this inconspicuous, albeit mostly non-sexual gametophytic stage of a taxon that was rare, and listed as a critically endangered and protected species, its presence posed profound implications for those drafting and enforcing conservation legislation. The gametophyte has been located within the known sporophytic distribution from Tenerife, Canary Islands; Madeira; the Azores; Algeciras, S. Spain, Asturias, N. Spain (Viane, pers. comm., 1992); Douro Valley, Portugal: Brittany (Jermy & Viane in Ripley 1990; Prelli, pers. comm., 1993); and Apuane Alpes, N. Italy; all sites recorded by A.C.J. and/or F.J.R. unless otherwise ‘stated. However, research has revealed that gametophyte populations extend into continental Europe far beyond the known sporophytic range of the species and closely parallel the known past distribution of Hymenophyllum tunbrigense (L.) Sm (cf. Richards & Evans 1972). The gametophyte generation is now known from several areas in central Europe. It is most widespread and relatively abundant in the Wasgau, the southern Pfalzerwald of Germany (C. Stark, pers. comm., 1996) and the adjacent N. Vosges of France (Jéréme et al. 1994). It is also reported from the sandstone massif shared by Luxembourg and Germany in the southern Eifel (Rasbach et al. 1993, 1995; Bujnoch & Kottke 1994; Reichling & Thorn 1997), and is known to be present as small scattered populations in the Elbsandsteingebirge, straddling eastern Germany and the Czech Republic (Vogel et al. 1993), and the Zittauer Gebirge (Jessen, pers. comm., 1997) — its easternmost known locality. It would appear to be thinly scattered in the intervening areas, with records from the northern Black Forest, east of Heidelberg in the Neckar valley, from the Spessart in the Main valley north-west of Wurzburg, from DISTRIBUTION OF THE KILLARNEY FERN GAMETOPHYTE 3 the Wupper valley near Solingen and from the northern Eifel near Monschau (Bennert et al. 1994; Kirsch & Bennert 1996). Over the past six years, an extensive survey by F.J.R. and A.C.J. and local recording by others in the British Isles has shown the gametophyte to be widespread, far beyond the present range of the sporophyte. Many suitable areas remain to be investigated but a summary of the position to early 1997 is presented here with the hope that it will stimulate additional recording. The wider gametophyte distribution reported here poses many questions, not least of which are: how is the disjunction between the generations perpetuated? (an issue addressed by Rumsey et al. 1992; and further discussed in Rumsey & Sheffield (1996)) and; when was the current range achieved and by what means? (the topic of an on-going research project at the Natural History Museum, London). -TRICHOMANES SPECIOSUM IN THE BRITISH ISLES Crucial to the understanding of the distribution of this species is the elucidation of the ecological differences between the two phases of the life-cycle, an aspect which may play a major role in perpetuating independence of either generation (Rumsey & Sheffield 1996). Much can be inferred from a comparison of the overall distribution of the two generations, the broad extent of which are given in Fig. 1. The distribution of the sporophyte within Britain has been outlined by Ratcliffe et al. (in Wigginton, in press). They report the past presence of 24 separate colonies, in 17 localities, occurring in a total of eleven widely scattered vice-counties. Only 16 colonies in ten localities are known to be extant. In Ireland only ten sites (in six vice-counties), out of the 43 once recorded were reported to be extant by Curtis & McGough (1988). The species is, however, clearly under- recorded, especially in the hill country of Counties Kerry and Cork where Ratcliffe et al. (1993) report the presence of 26 of the 30 Irish colonies known to them over the past three decades. The importance of these rare sporophyte colonies as potential sources, by means of spore dispersal, of the wider gametophyte distribution, remains to be resolved and is currently under investigation. The history of the discovery of the sporophyte in various areas of the British Isles is not elaborated further here but is discussed in part by Roberts (1979) and Church (1990). Since the discovery of the distinctive filamentous gametophyte generation and with growing awareness of its habitat preferences, wider surveys have revealed it to be remarkably widespread, if often extremely localized. We must assume that a combination of a morphology not readily assignable to any one cryptogamic group, coupled with growth in a poorly investigated and often inaccessible environment has resulted in its being overlooked for so long. Given this oversight a brief description of the habitats in which the gametophyte may be found is given below. MORPHOLOGY OF THE GAMETOPHYTE The morphology of the gametophyte has been described and illustrated (Rumsey et al. 1990; 1993) but is described again here with the hope that a wider audience will come to recognize it. Trichomanes speciosum gametophytes consist of branched filaments, the individual cells of which are c. 40-55 um wide and 150—300 um long, that grow interwoven into tufts or mats with an open, felt-like appearance (Fig. 2). These are of a clear bright glowing green when well hydrated, taking on a somewhat bluish-black metallic cast as the filaments crumple on drying. The gametophyte colonies can vary in overall size, from occurring as scattered filaments among bryophytes, to more or less pure patches covering several square metres to a depth of about one centimetre. The majority of sites, however, support small tufts ranging from thumbnail-sized patches to up to c. 10 cm*. The combination of colour, shape and restriction to particular niches within habitats makes field recognition of the gametophyte relatively easy in the majority of cases. The filaments maintain a rigidity, giving a distinctive wool-like resilience, when lightly touched, and by which an experienced worker can identify the colony or mat. They are distinguished from bryophyte protonemata by their larger diameter filaments, the cells of which are without oblique end-walls, and from filamentous green and yellow-green algae by their pale brownish rhizoids and the presence | of characteristically-shaped gemmifers, gemmae and gametangia (sex organs), when present. The 4 F. J. RUMSEY, A. C. JERMY AND E. SHEFFIELD Ficure 1. World distribution of Trichomanes speciosum Willd.: @ both generations; © gametophyte only. All known records mapped on the Atlas florae Europaeae base map (Jalas & Suominen 1972) amended to include the Canary Islands and Madeira. Dots indicate the presence of the species at any time within 50 km squares of the UTM grid map. DISTRIBUTION OF THE KILLARNEY FERN GAMETOPHYTE a Sa aia QEg§ no v 28 “OS Bae ee a {vt (au es| eecsu ce oye = > 2S 48 (oe) Se Oo Mu Ie ot AD Ses | oO i Qe ae aD-a a ie oO S12 3 & = iS ‘ea w= a = 3 oy ey Sige a5 S a lol 0) = ion tot Lower. Scanning electron micrograph of a gametophyte ‘‘mat”’ (x 30). (Micrograph, Figure 2. Upper. Typical gametophyte hab Farrar, who first drew attent 6 F. J. RUMSEY; A. C. JERMY AND E? SHEFFIELD most superficially similar algae grow in wet, well illuminated positions, in which T. speciosum gametophytes have never been found. Discrimination from other ferns is straightforward as no other native European fern genera produce filamentous gametophytes. As the gametophytes of all taxa within Trichomanes Section Lacosteopsis Prantl (= Vandenboschia Copeland) are so morphologically similar as to be effectively indistinguishable, certainly in the field, the possibility that one or more “independent gametophyte”’ taxa (i.e. lacking a known sporophyte) may occur in Europe, asin N. America, must be considered. This is particularly likely to be the case given that the diploid progenitors of the tetraploid T. speciosum are unknown. Available molecular evidence would suggest, however, that all material examined to date is of one, admittedly variable, taxon. Confusion following the spread of exotic species is unlikely but cannot entirely be ruled out (e.g. Rumsey et al. 1993). The gametangia are very similar in structure to those of other Trichomanes sensu lato (Stokey 1948; Yoroi 1972) and are obviously of a Filicalean form (see Fig. 3). The archegonia are produced on a specialised structure, the archegoniophore which is borne on a short broader filament and is produced just above the substrate, deep in the gametophyte tufts. They are apparently rarely produced, being found in less than 10% of gametophytes collected from the field, and later development in material grown in the laboratory has made no significant difference to this figure. In contrast, the antheridia, which may be found on the same filament, are more readily produced; c. 25% of gametophytes had at least one when collected and following cultivation nearly 75% had produced them (Rumsey & Sheffield 1996). Antheridial dehiscence with functional (i.e. motile) antherozoids has, however, rarely been observed. Thus the potential for gametophyte colonies to generate sporophytes cannot be assumed. DISTRIBUTION OF THE GAMETOPHYTE IN THE BRITISH ISLES All gametophyte records have been lodged with the Biological Records Centre, Monks Wood. While details of the gametophyte’s distribution remain to be established, especially in Scotland and Ireland, the provisional map (Fig. 4) probably represents an accurate picture of its distribution and general regional abundance. Up to February 1997, gametophytes have been recorded in 38 British vice-counties: 1, 2, 3, 4, 14, 34, 35, 36, 39, 41, 42, 43, 44, 45, 46, 47, 48, 49, 52, 57, 62, 63, 64, 67, 69, 70, 88, 95, 98, 100, 101, 102, 103, 104, 105, 107, 108 and 109 and in 120 10-km grid squares (hectads). In three of these, both those in v.c. 1 (West Cornwall), and one in v.c. 2 (E. Cornwall) it is present in grottoes and artificial features in gardens, and in two cases closely associated with sporophytes which are assumed to have been deliberately introduced. It is currently recorded from 13 Irish vice- counties: H1, 2, 3, 6,8, 10, 13, 16, 20, 26, 27, 33, 35 and 22 10-km squares but has not been searched for as exhaustively as in some other places. Some local botanists have been encouraged by finding both topography and geology in their area suitable for the gametophyte and have quickly come to recognise both its habitat and morphology. We believe that as more become acquainted with the gametophyte, much of the appropriate country where sandstone and the coarser volcanic rocks predominate will be shown to house this stage of the Killarney Fern. It should be noted, however, that although widespread, the gametophytic generation is by no means common. More than 75% of the grid squares mapped contain only single populations, in many cases restricted to a single microtopographical feature. Furthermore, we must stress (and see below) that the species in its entirety is protected under laws in the European Union, the U.K. and the Republic of Ireland. After many years of observation of the sporophyte generation Ratcliffe et al. (1993) concluded its distribution “. . . is puzzling in that it is absent from a great many apparently suitable habitats within its climatic range’. The problem is that too little is currently known of the environmental constraints on either generation’s growth and survival. The apparently anomalous distribution of this species, as with many other rare “Atlantic” cryptogams which show great individual longevity but very little or no current dispersive ability, can be arguably best explained as the product of rare climatic and stochastic events. Given no, or very limited, ability to recolonize once lost, the occurrence of these species implies a local continuity of site/habitat suitability, where an absence merely suggests that conditions have become unsuitable, if only once in the last few hundred years. It is thus not surprising that attempts to match this species’ distribution to climatic factors, often expressed as means, have met with only limited success. DISTRIBUTION OF THE KILLARNEY FERN GAMETOPHYTE i Ficure 3. Trichomanes speciosum gametophytes. A. Germinating spore (r = rhizoid); B. Gametophyte with archegoniophores (¥); C. Gametophyte with antheridium (C’) and gemma; D. Gemma; E. Gemmifers; F. Archegoniophore with archegonia; G. Antheridia. Scale bar = 100 um. 8 F. J. RUMSEY, A. C. JERMY AND E. SHEFFIELD Ficure 4. Distribution of the Trichomanes speciosum Willd. gametophyte generation in the British Isles (1989- LOOT). GAMETOPHYTE SITE RECORDS First records only for each tetrad are listed below. Vice-county names follow Dandy (1969) for those in Britain, and in Ireland they are as adopted by Scannell & Synnott (1987). In order to protect the location of extant sporophyte sites in Ireland we have cited in many cases a hectad reference only and an appropriate broad locality. The figure in square brackets after the hectad reference is the DISTRIBUTION OF THE KILLARNEY FERN GAMETOPHYTE 9 number of tetrads in which the gametophyte has been recorded when that is more than one; the recorder and date that follows is the first finding in that hectad. The altitude range at which the gametophyte is found in Ireland is similar to that in Britain, being from near sea-level to 300 m on Brandon Mountain but have not been cited in many cases for the above reason. The following abbreviations are used: ADH = A. D. Headley; AH = A. D. Hale; AOC = A. O. Chater; AP = A. M. Paul; ARC = A. R. Church; BM = B. Meatyard; CJ = A. C. Jermy; CP = C. D. Preston; CR = C. J. Raine; DB = D. Batty; DCB = D. C. Bryce; DF = D. R. Farrar; DH = D. M. Henderson; DJ = D. Jones; DP = D. A. Pearman; DR = D. A. Ratcliffe; EC = E. G. Cutter; ES = E. Sheffield; FR = F. J. Rumsey; GS = G. A. Swan; HB = H. H. Birks; IM = I. K. Morgan; JB = H. J. Birks; JH = J. S. Holmes; JM = J. Mitchell; JPW = J. P. Woodman; JV = J.C. Vogel; JW = J. Walls; LF = L. Farrell; LG = L. R. Gander; MD = C. M. Dowlen; MR = M. H. Rickard; NK = N. Kingston; NS = N. F. Stewart; PT = P. Thompson; RC = R. J. Cooke; RF = R. FitzGerald; RM = R. J. Murphy; RN = R. H. Northridge; RS = R. J. Stewart; RW = R. M. Walls; SC = S. P. Chambers; SD =S.R. Davey; S&JG =S. & J. B. Grasse; SM = S. J. Munyard; SR = S. M. Rumsey; TR = T. C. G. Rich; WC = W. Condry. THE RECORDS V.c. 1 W. Cornwall SW/77.27. Glendurgan, 50 m alt., FR, SR & ES, Apr. 1991; SW/80.39. Tregye Wood (private garden), nr Truro, 55 m alt., FR, SR & ES, Apr. 1991; SW/99.46. In old fern grotto, Heligan Gardens, 3 km NW of Mevagissey, 70 m alt., CJ, Sept. 1990. V.c. 2 E. Cornwall SX/07.88. St Nectan’s Glen, nr Tintagel, 80 m alt., FR, Sept. 1990; SX/07.89. Rocky Valley, nr Tintangel, 20-50 m alt., CJ, Sept. 1990; SX/08.88. St Nectan’s Kieve, nr Tintagel, 120 m alt., CJ, Sept. 1990; SX/12.89 & 12.90. Lesnewth stream, 125-165 m alt., FR, SR & ES, Apr. 1991; SX/ 25.52. In cliff by promenade, West Looe, <3 m alt., NS & RS, June 1994. V.c. 3 S. Devon SX/48.80. Blacknor Park, W. side of River Burn, c. 200 malt., NS, Aug. 1995; SX/50.83. Lydford Gorge, by the White Lady Waterfall, 120 m alt., NS & RS, Oct. 1996; SX/50.84. Lydford Gorge, 150 m alt., FR, Sept. 1990; SX/61.77. Wistman’s Wood NNR, 420 m alt., RC, 1995; SX/70.90. Whitewater,W of Great Tree Hotel, c. 170 malt., NS, Oct. 1996; SX/76.79 & 76.80. Becka falls, nr Manaton, 160 m alt., FR, SR & ES, Apr. 1991. V.c. 4N. Devon SS/66.49. By small waterfall on coast path, nr Hollow Brook, 150 m alt., RC, Mar. 1997; SS/ 67.48. Woody Bay, 130 m alt., FR, SR & ES, Apr. 1991; SS/70.49. Mother Meldrum’s Cave, Valley of Rocks, 200 m alt., FR, SR & ES, Apr. 1991; SS/73.48. Myrtleberry Cleave/Lyn Cleave, Lynmouth, 50 m alt., FR, SR & ES, Apr. 1991; SS/74.48. Watersmeet, 120 m alt., FR, SR& ES, Apr. 1991. V.c. 14 East Sussex TQ/29.28. Nr Spicer’s Farm, SW of Balcombe, 76 m alt., RC, SD & TR, Mar. 1995; TQ/31.29. Balcombe Mill, crevice above waterfall, c. 50 m alt., RC, SD & TR, Mar. 1995; TQ/33.31. Wakehurst Place, Ardingly, 75 m alt., CJ & MD, Aug. 1993; TQ/35.32. Philpots, West Hoathly, ie male he, Mar. 1997; TO/55.36. Eridge Rocks, Eridge Green S.S.S-1., 76 m alt., RC, Feb. 1997; TQ/83.10. Ecclesbourne Glen, nr Hastings, 70 m alt., SM, Feb. 1993; TQ/85.11. Fairlight Glen, c. 7 km E Hastings, 76 m alt., CJ, JV, MG & RC, Nov. 1992. V.c. 34 W. Gloucs. ST/62.76. R. Frome valley, E. Bristol, 30 m alt., C & MK, Mar. 1997; SO/5412. Rodge Wood nr Staunton, 230-260 m alt., FR, MG, AP, MR & JV, Feb. 1997. V.c. 35 Mons. S$O/52.03 & 52.04. Cleddon Shoots Reserve, Cleddon, 185-200 m alt., FR, MG, AP & JV, Feb. 1997. V.c. 36 Herefs. SO/61.23. Penyard Park, Ross-on-Wye, 150-160 m alt., FR & MR, Feb. 1997. V.c. 39 Staffs. SJ/98.65. Lud’s Church, 2 km NE of Swythamley Hall, 320 m alt., EC, FR & ES, Jan. 1991. 10 F. J. RUMSEY, A. C. JERMY AND E. SHEFFIELD V.c. 41 Glam. SN/82.01. Melincourt, Glamorgan W.L.T. Reserve, at waterfall, c. 75 m alt., RC, 1993. Vc: 42 Brees: SO/12.38. Brechfa Common, nr Llyswen, 300 m. alt., CJ, Apr. 1990. V.c. 43 Rads. SO/11.42 & 12.43. Cwm Bach Howie gorge, 12 km W Hay-on-Wye, c. 200 m alt., RC, 1993; SO/ 1.7. Lawn Brook, Beacon Hill, 135 m alt., CJ, Apr. 1990. V.c. 44 Carms. SN/44.07. Cwm Clydach, 50 m alt., IM, May 1991. V.c. 45 Pembs. SN/24.37. Cwm Cych, nr Cenarth, IM, May 1991; SN/26.35. Cwm Cych, nr Cenarth, 40 m alt., AOC Apr SoI: V.c. 46 Cards. SN/19.43. N of Coedmor mansion, Teifi estuary, 25 m alt., AOC, Mar. 1994; SN/19.44. Just S of Cwm Du, Coedmor, Teifi estuary, 20 m alt., AOC, Apr. 1991; SN/22.46. By Nant Arberth N of Point Rhyd-Arberth, 40 m alt., AOC & LG, May 1995; SN/29.52. NW of Llanborth, Penbryn, 15 m alt., AOC, Mar. 1994; SN/43.60. SW of Panteryrod, Afon Drywi valley, 50 m alt., AOC, Jan. 1996; SN/49.62 & 50.62. By Afon Arth, nr Monachty-back, 75-80 m alt., AOC, APF, Apr. 1991; SN/56.68. NNE of Plas Howell, Nant Rhydrosser, 115 m alt., AOC, Feb. 1992; SN/56.70. E of Pencwmisaf, Wyre valley, 65 m alt., AOC, July 1994; SN/59.59 & 59.60. Coed Gwenffrwd, nr Llangeitho, 120 malt., AOC, APF & DCB, Apr. 1991; SN/66.92 & 67.91. Coed Cwm Clettwr, nr Tre’r-ddol, 60-110 m alt., AOC & CJ, Apr. 1991; SN/69.94. S side of Cwm Einion, ESE of Ty’n- y-garth, 80 m alt., AOC, Nov. 1993; SN/69.96. By Afon Melindwr in garden of Llwyncelyn, Eglwys-fach, 15 malt., AOC, Feb. 1994; SN/70.94. By Afon Einion, S of Dol-goch, Cwm Einion, 200 m alt., AOC, Nov. 1993; SN/71.97. Llynfnant valley, W end of scree, just east of Allt-ddu, 40 malt., AOC, CJ, Apr. 1991; SN/73.72. Below road bridge WNW of Pontrhyd-y-groes, 150 malt., AOC, July 1994; SN/73.77. Coed Rheidol NNR, E of waterfall, Derwen, WNW of Devil’s Bridge, 160 m alt., AOC & CJ, Apr. 1991; SN/73.96. By Nant Cefn-coch, Llyfnant, 110 m alt., AOC, SC & AH, Feb. 1994; SN/73.97. Llyfnant Valley, east of Allt-ddu, 110 m alt., FR, Jan. 1991; SN/74.77. W side of Mynach Falls, Devil’s Bridge, 120 m alt., AOC & JPW, Feb. 1995; SN/ 74.78. Coed Rheidol NNR, NW of Erwbarfau, 150-180 m alt., AOC, CJ, FR & CR, Feb. 1992; SN/76.48. Craig Ddu, Doethie, 320 m alt., AOC, Sept. 1992; SN/76.73. Below Mariamne’s Garden, Hafod, 190 m alt., AOC, Dec. 1993. V.c. 47 Monts. SH/77.10. Below Lwydiath Hall, Afon Dulas, 130 m alt., CJ, FR, WC, S & JG, Apr. 1992. V.c. 48 Merioneth SH/62.38. Y Garth, 30 m alt., FR, 1996; SH/63.13. Afron Morfa, SW of Arthog, side of stream gorge, c. 90 malt., RC, July 1993; SH/63.37. Coed Caerwych, 105 m alt., FR, Jan. 1991; SH/66.18. Bontddu, 40m alt., FR, Jan. 1991; SH/65.39 & 66.38. Ceunant Llennyrch, nr Maentwrog, 30-90 m alt., DF & FR, Nov. 1989; SH/69.41. Ceunant Cynfal, nr Maentwrog, 60 m alt., CJ & FR, Apr. 1992; SH/72.11. Cader Idris, Nant Cader and scree on south side of Llyn Cau, 210 m alt., CJ, FR, WC,S & JG, Apr. 1992; SH/75.18. Torrent walk, 130 malt., CJ, FR, WC, S & JG, Apr. 1992; SH/ 79.10. Nr Aberllefenni, in F.C. woodland, 205 m alt., CJ, FR, WC, S & JG, Apr. 1992. V.c. 49 Caerns. SH/55.47. Cwm Llefrith, Moel Hebog and adjacent Moel yr Ogof, 410-500 m alt., FR, Mar. 1990; SH/63.51. Clogwyn y Barcut, 150 m alt., ADH, Jan. 1992; SH/66.70 & 66.71. Aber Falls NNR, 120-200 m alt., RC, 1991; SH/72.57. Coed Bryn Engan, 190 m alt., ADH, Apr. 1990; SH/76.68. Conway Valley, west of Tal y bont, at Caer-illin-ford, on the tributary Afon Dylin, 155 m alt., CJ & FR, Apr. 1991; SH/77.66. Coed Dolgarrog N.N.R., along Afon Ddu, 120 m alt., RC, 1992. V.c. 52 Anglesey SH/55.73. Cadnant Dingle S.S.S.I., 1 km S of Llandegfan, 75 m alt., RC, 1991. V.c.57 Derbys SK/24.76 & 25.77. Froggatt Edge, c. 1 km NE of the village of Froggatt, 250-270 m alt., FR & ADH, Jan. 1990. Weer 625N ES Yorks SE/58.98. Tripsdale, 260 m alt., KT, Oct. 1996; SE/81.91. Newton Dale, Newton-on-Rawcliffe, DISTRIBUTION OF THE KILLARNEY FERN GAMETOPHYTE 1] 120 m alt., KT, Mar. 1997; SE/83.84. Hyggitt’s Scar, nr Pickering, 180 m alt., KT, Feb. 1997; SE/ 94.97. Castlebeck Wood S.S.S.I., Harwood Dale, 100 m alt., K.T., Apr. 1996; SE/94.98. Bloody Beck S.S.S.I., 120m alt., KT, Apr. 1996; SE/99.98. Hayburn Beck, 4km N of Cloughton, 0-90 m alt., KT, Aug. 1995; SE/99.99. Beast Cliff, 9 km N of Scarborough, 90 m alt., KT, Oct. 1995; NZ/ 64.13. Wileycat Beck, nr Charlton, 220 m alt., KT, Feb. 1997; NZ/64.14. Wileycat Wood, nr Charlton, 160 m alt., KT, Feb. 1997; NZ/65.03. Wood Dale, Waites Moor, 250 m alt., KT, Mar. 1997; NZ/70.03 & 71.04. Great Fryup Dale, 330-350 m alt., KT, Mar. 1997; NZ/70.14. Mill Beck Woods, Moorsholm, 130 m alt., KT, Nov. 1996; NZ/70.17. Mains Wood, Loftus, 65 m alt., CJ, FR, & KT, Feb. 1996; NZ/71.18. Whitecliffe Wood, Loftus, c. 40 m alt., CJ, FR & KT, Feb 1996; NZ/71.02 & 72.02. Woodhead, Great Fryup Dale, 300 m alt., KT, Sept. 1996; NZ/72.06 & 73.06. Danby Crag Wood, 160-220 m alt., KT, Oct. 1995; NZ/72.17. Wauple Wood, nr Liverton, 140 m alt., KT, Dec. 1996; NZ/73.01, 74.01 & 74.02. Glaisdale Head Crag, 280 m alt., KT, Oct. 1996; NZ/74.14. Upper Roxby Woods, 140 m alt., KT, Feb. 1997; NZ/75.07. Crunkly Gill, Leatholme, 120 m alt., KT, Oct. 1996; NZ/75.16. Lower Roxby Woods, 90 m alt., KT, Feb 1997; NZ/77.05. Glaisdale Wood, 80 m alt., KT, Oct. 1995; NZ/77.08. Stonegate Gill, nr Leatholme, 130 m alt., KT, Nov., 1995; NZ/78.04. West Arnecliffe Woods, nr Egton Bridge, 100-170 m alt., KT, Nov 1994; NZ/79.03. Park Hole Wood, nr Egton Bridge, 120-150 m alt., KT, May 1996; NZ/79.05. Limber Hill Wood, nr Egton Bridge, 90 m alt., KT, Oct. 1996; NZ/80.04. Below Blue Beck Cottage, Egton bridge, 60-100 m alt., KT, Nov. 1996; NZ/81.00. Scar Wood S.S.S.I., nr Goathland, 100-120 m alt., KT, Mar. 1996; NZ/81.02. Combs Wood, nr Beck Hole, 130-160 m alt., KT, Mar. 1996; NZ/81.04. Spring Wood, nr Grosmont, 130 malt., KT, Mar. 1996; NZ/81.06. Hunter Hill, nr Grosmont, 90 m alt., KT, Nov. 1996; NZ/81.13. High Dale nr Mickelby, 120 m alt., KT, Nov. 1996; NZ/82.01. Carr Wood and Mallyan Spout, nr Goathland, 80-90 m alt., ADH, Sept. 1993; NZ/80.12. Thomason Foss, nr Beck Hole, 110 m alt., MD, Aug 1994; NZ/ 82.04. Crag Cliff Wood, nr Grosmont, 70-90 m alt., KT, Mar. 1996; NZ/83.04. Lythe Beck, nr Grosmont, 50-150 m alt., KT, Nov. 1996; NZ/83.11 & 84.11. Mulgrave Woods around Biggersdale Hole waterfall, 80 m alt., KT, June 1995; NZ/85.11. Mulgrave Woods, nr Sandsend, 90 m alt., KT, Jan. 1997; NZ/86.11. Dunsley Woods, nr Sandsend, 50 m alt., KT, Nov. 1996; NZ/ 88.03. Nr Falling Foss, S of Littlebeck, 110 m alt., KT, Apr. 1995; NZ/88.04. Nr Littlebeck, 90 m alt., KT, Dec. 1996; NZ/92.03. Oak Wood, nr Fylingthorpe, 100 m alt., KT, Dec. 1996; NZ/90.07 & 91.07. Rigg Mill Wood, nr Hawkser, 90 m alt., KT, Dec. 1996; NZ/90.08. Cock Mill Wood, nr Ruswarp, 30 m alt., KT, July 1995; NZ/95.01. Howdale Wood, nr Stoup Brow, 130 m alt., KT, Dec. 1996; TA/00.96. Hayburn Wyke, 2 km N of Cloughton, 5—90 m alt., FR, SR & ADH, Jan. 1992; TA/00.99. Beast Cliff, 9 km N of Scarborough, 90 m alt., KT, Oct. 1995. Mee. .O3 S-aWwi Yorks. SD/96.26. Jumble Hole Clough, c. 240 m alt., ADH, Sept. 1991; SD/97.26 & 98.26. Callis Wood, SW of Hebden Bridge, 130-200 m alt., DF, CJ & FR, Nov. 1989; SD/97.25. Dill Scouts Wood/ Colden Clough, nr Hebden Bridge, c. 170 m alt., DF, CJ & FR, Nov. 1989; SD/97.29. Greenwood Lee, 220 m alt., FR & ADH, Mar. 1990; SD/97.30. Hardcastle Crags, 200-225 m alt., ADH, Sept. 1991; SK/23.96. Spout Brow/Bull Clough, 250 m alt., FR, Aug. 1990; SK/23.97. Valley between Ewden Lodge Farm and Cottage Farm, 260 m alt., FR, Aug. 1990; SK/24.96. Ewden Beck, 240 m. alt., ADH, Dec. 1989; SK/24.93. Agden Beck, 220 m alt., ADH, June 1990; SE/07.36. Harden Beck, 170m alt., MG, FR, ES & JV, Feb. 1993; SE/10.38. Bell Bank, Bingley, on either side of A65 (previously B6429), 100-140 m alt., DF, CJ, FR, Nov. 1989. V.c. 64 Mid-W. Yorks. SE/06.56. Strid Woods, Wharfedale, 2 km N of Bolton Abbey, 115—120 m alt., FR & ADH, Mar. 1990; SE/07.56. Valley of Desolation, Hudson Gill Beck, 2 km N of Bolton Abbey, 180m alt., FR & SR, July 1990; SE/09.46 & 09.47. Hebers Ghyll, 1-5 km SW of Ilkley, 185-235 m alt., FR & ADH, Dec. 1989; SE/10.47. Panorama Wood, 1:5 km SW of Ilkley, 200 m alt., FR & ADH, Dec. 1989; SE/12.50. West Moore and March Ghyll, 180-190 m alt., ADH, FR & SR, Mar. 1991; SE/ 15.63 & 15.64. Ravensgill, Bewerley, 180-240 m alt., FR & ADH, Jan. 1990; SE/20.63. Braisty Woods, nr Summerbridge, 235 m alt., FR & ADH, Sept. 1993; SE/20.64. Brimham Rocks, 270 m alt., ADH, 1996; SE/22.44. Danefield Wood, Otley, 170 m alt., FR, Sept. 1993; SE/23.77. Hackfall, 1 km NE of Grewelthorpe on SW side of River Ure, 130 m alt., FR & ADH, Jan. 1990; SE/27.54. Cardale Woodland, nr Harrogate, 108-115 m alt., FR & ADH, Sept. 1993; SE/35.53. Plumpton Rocks, 45 m alt., FR & ADH, Nov. 1996. 12 F. J. RUMSEY, A. C. JERMY AND E. SHEFFIELD V.c. 67 S. Northumb. NY/73.83. Roughside Moor, Cragshield Hope, 10 km W of Bellingham, 245 m alt., CJ & GS, July, 1995. V.c. 69 Westmorland SD/17.99. Stanley Force, 0-5 km SW of Eskdale, c. 100 m alt., DF, Oct. 1989; NY/36.06. Rydal Beck 1 km N of Rydal, 130 m alt., DF, CJ & FR, Nov. 1989; NY/36.07. Rydal Beck, 165 m alt., FR, Jan. 1991. V.c. 70 Cumberland NY/21.01. Hardknott Gill, 150 m alt., ADH, Dec. 1991; NY/26.18. Watendlath Beck, above and below Lodore Falls; c. 80-100 m alt. DF, Oct. 1989; NY/50.49. Eden Brows/Froddle Crook, 45- 60 m alt., ADH, Nov. 1995. V.c. 88 Mid Perth NO/00.41. The Hermitage, Dunkeld, c. 100 m alt., FR, July 1995. V.c. 95 Moray NJ/18.71. Covesea, on raised beach 5 km W of Lossiemouth, 20 m alt., CJ & JV, July 1992. V.c. 98 Main Argyll NR/96.66. Kilbride Bay, Loch Fyne, 1 km N of Ardlamont Point, 5 m alt., CJ, July 1991; NS/ 00.77. Loch Riddon, N of Tignabruaich, c. 120 m alt., DR & JM, Apr. 1996. V.c. 100 Clyde Is. NR/86.40. Arran, Imachar Point, 25 m alt., FR, Oct. 1993; NR/86.41. Arran, N. of Imachar, 15m alt., ARC, Aug. 1996; NR/88.31. Arran, shoreline just north of King’s Cave, c. 5 m alt., FR & SR, Sept. 1996; NR/89.35. Arran, Machrie Bay, 20 m alt., FR, Oct. 1993; NR/90.48. Arran, S of Catacol Bay, 10 m alt., ARC, Sept. 1994; NR/93.51, Arran, Lochranza, 5 m alt., RC, May 1997; NR/94.52. Arran, Lochranza, Fairy Glen, 90-120 m alt., ARC & CJ, June 1991; NR/95.52. Arran, just east of An Scriodan, 85 m alt., FR, Sept. 1996; NR/96.51. Arran, Picture (Ossian’s) cave and woodland 1 km SE of Cock of Arran, 25-35 m alt., FR & ARC, Oct. 1993; NS/01.37. Arran, Brodick Castle grounds, 20 m alt., ARC, Feb. 1994; NS/01.44. Arran, Sannox Bay, 25 m alt., FR & ARC, Oct. 1993; NS/01.46. Arran, North Sannox, rocks by sea, c. 6 m alt., CJ, June 1991; NS/02.39. Arran, S of Pirates Cove restaurant, c. 10 m alt., FR & ARC, Oct. 1993; NS/ 02.40. Arran, S of Rubha Salach, c. 10 malt., FR & ARC, Oct. 1993; NS/03.29. Arran, Lamlash Bay, below Gortonallister, 7-40 m alt., FR, Oct. 1993; NS/03.35. Arran, Brodick-Corriegills Point, 10-35 alt., FR & ARC, Oct. 1993. V.c. 101 Kintyre . NR/66.32. Bellochantuy, 20 m alt., FR & SR, Sept. 1996; NR/73.19. Glenramskill, 25 m alt., FR & SR, Sept. 1996; NR/74.76. Between Eilean na Bruachain and St Columba’s Cave, 30 m alt., FR & SR, Sept. 1996; NR/76.15. Shoreline N of mouth of Balnabraid Glen, 5 m alt., FR & SR, Sept. 1996; NR/76.77. Roadside at Caolisport, 15 m alt., FR & SR, Sept. 1996; NR/81.37. Port Righ, Carradale, 15 malt., FR & SR, Sept. 1996; NR/85.76. Artilligan Wood, 30 m alt., FR & SR, Sept. 1996; NR/85.78. Just N of Nead an Fhitich, 30 m alt., FR & SR, Sept. 1996; NR/86.55. E bank of Allt a Bhuie, Claonaig, 10 m alt., FR & SR, Sept. 1996; NR/91.60—91.62. Skipness to Rubha Grianain, 0-25 m alt., CJ & DB, July 1991. V.c. 102 S. Ebudes NR/37.45. Islay, Sr6n Dubh, between Laphroaig and Port Ellen, 20 m alt., FR & SR, Sept. 1996; NR/41.73. Islay, Bunnahabhain, <5 m alt., FR & SR, Sept. 1996; NR/44.67. Jura, raised beach cliff, 1-5 km S of Feolin Ferry, 20 m alt., FR & SR, Sept. 1996; NR/44.72. Jura, raised beach beyond Inver Cottage, 15 m alt., FR & SR, Sept. 1996; NR/46.53. Islay, Claggain Bay, 5 m alt., FR & SR, Sept. 1996; NR/46.76. Jura, coast S of Loch na Sgrioba, c. 3 m alt., LF & CP, June 1991; NR/48.63. Jura, below Jura House, Brosdale, in sea caves, c. 3 malt., CJ & LF, June 1991; NR/67.99. Jura, Beinn nan Capull, in cave by shore, 2-5 m alt., CJ, June 1991; NR/70.98. Jura, woodland above Kinuachdrachd Harbour, 30 m alt., FR & SR, Sept. 1996; NM/67.00. Jura, cave on Uirigh Ghlas, 75 m alt., CJ, June 1991. V.c. 103 Mid Ebudes NM/43.42. Isle of Mull, in crevices along shore, 3 km N of Ulva Ferry, 12 m alt., CJ, JH & BM, Sept 1995; NM/55.21. Isle of Mull, Carraig Mhor, 1 km E of Carsaig, 3-5 m alt., CJ, JH & BM, Sept di995" DISTRIBUTION OF THE KILLARNEY FERN GAMETOPHYTE 13 V.c. 104 N. Ebudes NG/61.11. Skye, Toskavaig, c. 100 m alt., HB & JB, July 1992; NG/73.15. Skye, Isleornsay, on Sound of Sleat, nr Rubha Guail, 10m alt., CJ & RC, Aug. 1991; NG/78.19. Skye, between Dunan Ruadh and Port Aslaiq, c. 10 m alt., HB & JB, July 1992; NG/76.24. Mudelach, Sron an Tairbh, 75 m alt., HB & JB, July 1992. V.c. 105 W. Ross NG/81.90. In cave once used as a church, S of Cove, nr Gairloch, 2 m alt., CJ & DH, July 1992; NC/07.03. Allt nan Coisiche, on western slopes of Ben More Coigach, about 0-5 km ENE of Culnacraig, 120 m alt., CJ & DP, July 1993. V.c. 107 E. Sutherland NH/76.98. Loch Fleet, on the NE facing cliffs of Creag an Armalaidh, 1 km NW of The Mound (Causeway), 12 m alt., CJ & RW, Aug. 1993. V.c. 108 W. Sutherland NC/18.61. On coast NW of Kinlochbervie, 1 km S of Rubh’an Fhir Leithe, 20 m alt., CJ & RW, Aug. 1993. V.c. 109 Caithness ND/18.73. W side of Dunnet Head, S of Dunnet Hill, nr Chapel Geo and W of old chapel, 175 m alt., JW & RW, Aug. 1993; ND/19.76. Dunnet Head, N of Burifa Hill on coast by Shira Geo, 8 m alt., JW & RW, Aug. 1993; ND/11.22. On steep banks of Langwell Water, at Berriedale, 17 m alt., CJ & RW, Aug. 1993. V.c. H1 S. Kerry Q/4.1. Brandon Mtn, FR & ADH, May 1993; Q/4.0. Brandon Mtn, RC, Aug. 1994; V/6.6. Cloonaghlin Lough, NE of Waterville, FR, May 1993; V/8.6.Cloonea Lough area, RC, July 1992; V/9.8. Killarney, RC, July 1992. V.c. H2 N. Kerry V/9.8. [4] Killarney area, FR & ADH, May 1993. V.c. H3 W. Cork W/1.3. Skibbereen, RF, Apr. 1994; W/2.3. Leap, NK, Feb. 1995; W/7.5. Hungry Hill, NS, Oct. 1993. V.c. H6 Co. Waterford S/0.0. Lismore, MD, July 1993. V.c. H8 Co. Limerick R/7.5 [2]. Shievefelim mountains area, FR & ADH, May 1993. V.c. H10 N. Tipperary R/7.5. Slievefelim mountains area, FR & ADH, May 1993. V.c. H13 Co. Carlow S/7.4 [2]. Tinnahinch, FR & ADH, May, 1993. V.c. H16 W. Galway L/67.62. NE slopes of Tully Mtn, 150 m alt., CJ, Aug. 1994. V.c. H20 Co. Wicklow O/19.12. Powerscourt Waterfall, 190 m alt., FR & PT, Sept. 1992; T/23.99. Devil’s Glen, c. 180 m alt., NK, Feb. 1995. V.c. H26 E. Mayo G/20.04. Pontoon, NE of Castlebar, 70 m alt., MD, July 1993. V.c. H27 W. Mayo F/55.04. Achill Is., Croaghaun, 180 m alt., CJ, Aug. 1994; F/99.35. Benmore, 230 m alt., CJ, Aug. 1994; G/10.06. Nephin, 210 m alt., CJ, Aug. 1994; G/15.01. Shanvoley, 80 m alt., CJ, Aug. 1994; L/67.62. Tully Mtn, Letterfrack, 120 m alt., CJ, Aug. 1994; L/92.65. N slopes Devil’s Mother, 130 m alt., CJ, Aug. 1994. V.c. H33 Fermanagh H/03.55. Lough Navar Forest, 280 m alt., NS, Feb. 1993; H/05.55. E. of Lough Navar, c. 200 m alt., RN, Jan. 1994; H/06.53. Derryvahon, at E end of cliff, c. 175 malt., RN, Mar. 1996; H/07.54. Correl Glen, nr Derrygonnelly, about 50 m below bridge, 237m alt., NS, Feb. 1993; H/08.54. Correl Glen, lower end, 180 m alt., RN, May 1994. V.c. H35 W. Donegal G/0.2. Glenveagh, NS, Sept. 1993; G/9.8. Ardnamona Woods, MD, July 1993. 14 Pid) RUMSEY; A; C:JERMYIAND:-E: SHEFFIELD THE HABITAT OF THE GAMETOPHYTE The microsites inhabited by gametophytes are usually under deep rock overhangs, in undercut areas at cliffbases, in sea caves, by streamsides, in recesses amongst boulders in small natural caverns and crevices, all relatively humid, but not wet (Fig. 2). Some sites appear particularly dry but there may be a retention of moisture by the rock strata sufficient to retain a high humidity. In some “‘dry”’ sites Trichomanes speciosum grows on soft rock or hard glacial debris which facilitates the movement of moisture through capillary action. As with many “Atlantic” cryptogams, the majority of sites have shaded, NE aspects (Ratcliffe 1968), although at higher altitudes (>270 m) the proportion of sites facing into sun, i.e. SW, rises significantly (see Fig. 5), perhaps to offset decreases in ambient temperature, or length of growing season. The gametophyte thus occurs in areas of low light intensity, most measured values being <1uEm7*s_', which represents 1% or less of the ambient light level exterior to the occupied site. The gametophyte has the ability to photosynthesize at very low irradiances, maintaining a positive carbon balance in an environment too dark for photosynthesis to occur in other terrestrial ferns for which reliable measurements have been reported (Rumsey et al. 1996), which allows the exploitation of a stable, competition-free, climatically moderated environ- ment and must be considered responsible for the species’ survival in much of northern and continental Europe. Occupied sites show remarkably little variation in diurnal and annual temperature and humidity (Rumsey 1994). The latter is obviously of crucial importance as the species is absent from habitats where there is considerable movement of air, e.g. amongst block scree and tumbled boulders in stream beds, where other filmy-ferns thrive. Most sites are near geological transitions where differential weathering provides suitable undercut areas, waterfalls and steep-sided valleys, where temperatures remain moderated and humidity high. A wide range of acidic to neutral rock substrates are occupied (andesites, gneiss, basaltic lavas, granite, schists, sandstones, mudstones, greywackes, slates, grits and conglomerates), although slow growth rates and assumed poor dispersal means that rapidly eroding substrata, such as shales, are not generally suitable. Absence from limestone may be more the result of physical rather than chemical factors, especially given the weakly basicolous nature of the sporophyte, and also the fact that the gametophyte is found on weathered basalt. Porous substrates such as millstone grits and sandstones in general support more extensive gametophytic growth than non-porous, e.g. slates and schists, even in less macroclimati- cally suitable areas, such as Yorkshire. In these more continental areas the plant is almost always closely associated with spring lines and sub-surface moisture, which may permeate the porous rocks to provide a moderated humidity in the drier months. Desiccation is as liable to occur in the winter months, when lack of woodland canopies and herbaceous cover allows greater air movement into microsites, as in the warmer summer months when precipitation is reduced. The majority of sites occur in woodland; those in open situations are predominantly coastal or rarely in hyper-oceanic montane areas, extending to c. 500 m on Moel yr Ogof. This latter habitat is more frequently occupied in the wetter, winter-warm, west of Ireland, as is true too, of the sporophyte (Ratcliffe et al. 1993). The woodland sites occupied have a long history but many have obviously been extensively felled at some stage in the past judging from the paucity of Atlantic bryophyte species (Edwards 1986). The extremely sheltered habitats of the gametophyte have, we suggest, allowed its survival through clear-felling episodes which have resulted in the loss of less ‘oceanic’? but more photophilic species. A similar explanation may be advanced to account for the presence of the liverwort Jubula hutchinsiae (Hooker) Dum., a regular associate of Trichomanes speciosum, in otherwise bryophytically depauperate localities. The species intimately associated with the gametophyte generation consist of a limited range of common, shade-tolerant calcifuge bryophytes, (cf. Ratcliffe et al. 1993, for sporophyte associations) of which the most regularly present are /sopterygium elegans (Brid.) Lindb. and Calypogeia arguta Nees & Mont., with Tetraphis pellucida Hedw. an important associate in gritstone sites. Whilst seldom growing intermixed, preferring constantly damp rock surfaces, Tetrodontium brownianum (Dicks.) Schwaegr. is a useful site indicator of suitable undercut areas. The presence of a range of the Atlantic species recognised by Ratcliffe (1968) within the more continental areas of central and eastern Britain indicate that a search for the gametophyte may well be rewarded, given the presence of suitable niches. The requirement for suitable habitats cannot be overemphasized; climatically suitable sites supporting very rich Atlantic assemblages, e.g. Coed Ganllywd, Merioneth, do not DISTRIBUTION OF THE KILLARNEY FERN GAMETOPHYTE 15 50 >270 m (n = 9) 40 30 20 10 N NE iS) Gz S SW W NW 50 151-270 m (n = 47) 40 wa z = 30 iS) ® = 6 = 20 Le?) & [j= ib) S) 10 ® ae 0 N NE E SEY’ Ss SW W NW 50 0-150 m 40 30 20 10 "N NE E SE S SW W_~ NW Figure 5. Gametophyte of Trichomanes speciosum microsite aspect at three different altitudes. Aspects have been simplified to the nearest of eight cardinal points; bar height is proportional to the number of records of that aspect in that particular altitudinal class. 16 F. J. RUMSEY, A. C. JERMY AND E. SHEFFIELD support gametophytes, presumbly as the bedding plane orientation has not allowed the develop- ment of undercut areas above the stream. On the other hand the gametophyte survives in climatically ameliorated pockets in marginal sites supporting a very impoverished Atlantic flora on the Yorkshire grits and Scottish sea cliffs. The gametophyte is almost entirely absent in the British Isles from artificial habitats such as tunnels, quarries and mine adits, etc., even where closely adjacent to occupied natural sites. It has been reported to us as present in an adit, at Carr Wood, Goathland, N. Yorks (K. Trewren, pers. comm., 1995), and the only site seen by the authors is an adit, perhaps of Roman age, at Watersmeet, N. Devon. This suggests that successful dispersal and establishment, whether by spores or gametophyte gemmae is, within the British Isles, of very infrequent occurrence, at least under the conditions which have prevailed in historical time. The restriction of the sporophytic stage to an artificial habitat (well shafts) in Brittany (Louis-Arsene 1953a—c; McClintock 1963) would appear to be something of a paradox. It is significant that of all the known sites were of wells of considerable age, the species being absent from apparently similar but younger well sites. Was long- distance spore dispersal and colonization briefly possible under particular climatic conditions in the mediaeval period? Successful spore dispersal and colonization is obviously still occurring within Macaronesia, as evidenced by the species’ presence on lava fields of known age on Terceira and Pico in the Azores, levada edging on Madeira, and on many road and trackside dry-stone walls on Flores, Azores, where the species achieves its greatest abundance. One of the greatest paradoxes of this species and other “Atlantic” cryptograms, is in explaining how species which show little or no current dispersal over much of their range managed to reach the mid-Atlantic and effectively colonize these volcanic islands some 1000 km or more from the nearest landmass on the western fringe of Europe. Consideration of the climatic factors which may currently distinguish the Azorean region from elsewhere and which may have prevailed more widely in the past, would seem to offer the best opportunity for explaining the conditions necessary for dispersal to occur. IMPLICATIONS FOR CONSERVATION AND THE LAW Few would argue against the stringent legal protection of a species attractive to and vulnerable from collectors, especially if known as very few scattered and effectively isolated individuals, as is true of the Killarney Fern. This rarity and vulnerability is reflected throughout the species’ range on the European mainland and its occurrence in Macaronesia is becoming increasingly threatened by changes in agricultural practices and laurel forest clearance (A.C.J. and F.J.R., unpublished observations). Accordingly, Trichomanes speciosum (which must be interpreted as both gener- ations) has received special attention and is afforded legal protection throughout under the Berne Convention on the Conservation of European Wildlife and Natural Habitats (Appendix I) and is covered by the Directive of the Council of European Communities on the Conservation of Natural Habitats and of Wild Fauna and Flora 1992 (Annexes II and IV). This latter document requires Member States to preserve habitats, a laudable intention which obviously requires as a prerequisite accurate delimitation and documentation of the species’ distribution. In Britain it is thus protected under Schedule 4 of the Conservation (Natural Habitats, etc.) Regulations, 1994 and Schedule 8 of the Wildlife and Countryside Act, 1981; in Northern Ireland, the species is protected under the Conservation (Natural Habitats, etc.) Regulations (N.I.), 1995 and Pt. 1 of Schedule 8 of the Wildlife (N.I.) Order, 1985. In the Republic of Ireland the fern is protected under the Flora Protection Order, 1987, made under Section 21 of the Wildlife Act, 1976. In all cases these laws prohibit (except under licence) the picking, uprooting or otherwise taking, purchasing or selling the plants, or wilfully altering, damaging, destroying or interfering with the habitat of the species so protected. The discovery of the distinctive, persistent populations of the gametophyte has resulted in a unique problem. The sheer abundance of sites both in the British Isles and in continental Europe precludes their effective protection as demanded by law. We thus run the risk under the current situation that legal protection may be considered inappropriate or unworkable. One solution may be effectively to ignore the gametophyte generation and redraft legislation to render the sporophyte alone as protected. Any such decision must primarily consider the role of the gametophyte in the production of new sporophyte plants. If the generations are isolated by failure to reproduce sexually DISTRIBUTION OF THE KILLARNEY FERN GAMETOPHYTE Ly and very limited fertility respectively (Sheffield 1994), then the gametophyte might be considered as effectively a discrete organism and treated accordingly. Ratcliffe et al. (1993) in an in-depth discussion on the ecology and conservation of the Killarney Fern observed that no new sporophyte colonies had formed in Britain and Ireland, close to existing colonies known to them, over the past 30 years. Recent study reveals the situation may be less bleak, with juvenile sporophytes seen in four Scottish, five English and two Welsh sites over the period 1990-1996. Of these, however, more than half have been at sites which already support sporophytes. In addition juveniles have shown high rates of mortality, all being lost from at least one site in this brief period (Rumsey ef al. 1991). The incidence of sporophyte production throughout the species’ range has a definite geographical bias, strongly suggesting climatic factors such as winter warmth are of importance (Rumsey & Sheffield 1996). If so, slight macroclimatic change may have a profound effect on rates of sporophytic recruitment and therefore anything which jeopardized the gametophyte’s survival should be avoided. Gametophytes are perhaps vulnerable to disturbance by collection, the long term effects of which are unclear. However, given the growth form and totipotency of small fragments, collection is unlikely to cause the species’ extinction at a site, but regrowth and recolonization is a slow process. Small bare areas left when material was collected under licence in 1989 could still be identified five years later! The gametophyte is distinctive and does not need to be gathered for identification, nor is it obvious to, or easily damaged by, casual observers. It is also unlikely to be lost as a result of habitat destruction in the majority of its sites. If a responsible policy of non-collection by enthusiasts (the raising of sporophytes from gametophyte material is not a realistic option for amateur growers) and an ongoing monitoring programme to detect novel sporophytes can be maintained, then legal protection might sensibly be restricted to the sporophyte alone. Until such time, both generations have full legal protection as stated above. ACKNOWLEDGMENTS F.J.R. and E.S. would like to thank the Natural Environmental Research Council which funded much of this work under grant GR3/7902 and for their continued support (GR3/09451) of work on this species at the Natural History Museum, London, in collaboration with the University of Cambridge. A.C.J. and F.J.R. are grateful to the former Nature Conservancy Council, the Countryside Council for Wales (C.C.W.), and Scottish Natural Heritage (S.N.H.) for help with fieldwork expenses in relation to this survey. A.C.J. thanks also the Royal Irish Academy (Praeger Fund) for support for fieldwork in Ireland. We would like to thank all those who have given us permission to visit sites and those who have accompanied us in the field, especially David Batty (S.N.H.), Arthur Chater (B.S.B.I.), Tony Church (B.S.B.I.), Bill Condry, Tom Curtis (National Parks and Wildlife Service, Eire), John Davies (C.C.W.), Lynne Farrell (S.N.H.), Alan Fowles (C.C.W.), Jack and Sabina Grasse (B.S.B.I.), Alistair Headley, Dewi Jones, Glynn Jones (C.C.W.), Gerry Sharkey (B.S.B.I.) and Ken Trewren (British Pteridological Society). We thank also those who have sent in records (and voucher material) from areas not visited by us, especially Jack Bouckley, Rob Cooke, Elizabeth Cutter, Ian Evans, Rosemary FitzGerald, Mary Gibby, Naomi Kingston, Megan Lyall (née Dowlen), Rose Murphy, Robert Northridge, Aileen O’Sullivan, Chris Preston, Catherine Raine, Nick Stewart, Ray Woods and Johannes Vogel. The data have been logged and material (when sent) studied, even though that record may not be listed in this abbreviated list. We would like to thank Rob Cooke, Tom Curtis, Lynne Farrell, John Faulkner, Mary Gibby, Andy Jones, Colman O’Criodain, Johannes Vogel and Richard Weyl! for commenting on earlier drafts of parts of this MS. We are particularly grateful to Don Farrar who at last woke us up to the fact that these gametophytes were hiding beneath our rocks some twenty years after his supervisor and mentor, Professor Herb Wagner, of Michigan University, had leapt from rock to rock in the Yorkshire Dales, with torch in hand, rousing the British Pteridological Society to find the plant he knew should be there! 18 F. J. RUMSEY, A. C. JERMY AND E. SHEFFIELD REFERENCES ALLEN, D. E. (1969). The Victorian fern craze. A history of pteridomania. Hutchinson, London. BENNERT, H. W., JAGER, W., LEONHARDS, W., RasBacu, H. & RasBacu, K. (1994). Prothallien der Hautfarns Trichomanes speciosum (Hymenophyllaceae) auch in Nordrhein-Westfalen. Floristische Rundbrief 28: 80. Botton, J. (1785). Filices Britannicae: an history of the British proper ferns. J. Binns, Leeds. Busnocu, W. & Korrke, U. (1994). Der Gametophyt von Trichomanes speciosum Willd. im Regierungsbezirk Trier. Dendrocopos 21: 225-230. Cuurcu, A. R. (1990). Recent finds of Killarney fern (Trichomanes speciosum Willd.) in Arran. Glasgow naturalist 21: 608-614. Curtis, T. G. F. & McGoueu, H. N. (1988). The Irish red data book. I. Vascular plants. Stationery Office, Dublin. Danby, J. E. (1969). Watsonian vice-counties of Great Britain. Ray Society, London. Epwarps, M. E. (1986). Disturbance histories of four Snowdonia woodlands and their relation to Atlantic bryophyte distributions. Biological conservation 37: 301-320. Farrar, D. R. (1967). Gametophytes of four tropical fern genera reproducing independently of their sporophytes in the Southern Appalachians. Science, N. Y. 155: 1266-1267. Farrar, D. R. (1985). Independent fern gametophytes in the wild. Proceedings of the Royal Society of Edinburgh 86B: 361-369. FarRAR, D. R. (1992). Trichomanes intricatum: The independent Trichomanes gametophyte in the eastern United States. American fern journal 82: 68-74. FARRAR, D. R., PARKS, J. C. & McALpin, B. W. (1983). The fern genera Vittaria and Trichomanes in the North- eastern United States. Rhodora 85: 83-92. GOEBEL, K. (1888). Morphologische und biologische Studien: II Zur Keimungsgeschichte einiger Farne. Annales Jardin Buitenzorg 7: 74-119. JALAS, J. & SuomineN, J. (1972). Atlas florae Europaeae — Distribution of vascular plants in Europe. 1 Pteridophyta (Psilotaceae to Azollaceae). Committee for mapping the flora of Europe & Societas Biologica Fennica Vanamo, Helsinki. JEROME, C., RaspacH, H. & Rassacu, K. (1994). Découverte de la fougére Trichomanes speciosum (Hymenophyllaceae) dans le massif Vosgien. Le monde des plantes 450: 25-27. Kirscu, H. & BENNERT, H. W. (1996). Erstnachweis von Gametophyten des Hautfarns Trichomanes speciosum Willd. (Hymenophyllaceae) in Bayern. Nachrichten des naturwissenschaftlichen Museums Aschaffenburg 103: 119-133. Louts-ARSENE, Fr. (1953a). Trichomanes speciosum en Bretagne. Bulletin de la Société botanique de la France 100: 6. Louts-ARSENE, Fr. (1953b). Nouvelle note sur le Trichomanes speciosum. Bulletin de la Société botanique de la France 100: 187-188. Louis-ARSENE, Fr. (1953c). Les stations des Trichomanes speciosum dans la region de Ploemel. Bulletin de la Société botanique de la France 100: 285-290. McCuintock, D. (1963). Reports of the Exhibition Meeting, 1962. Proceedings of the Botanical Society of the British Isles 5: 170-171. RasBacH, H., RasBacH, K. & JEROME, C. (1993). Beobachtungen tiber das Vorkommen des Hautfarns Trichomanes speciosum (Hymenophyllaceae) in den Vogesen und dem benachbarten Deutschland. Carolinea 51: 51-52. RasBacu, H., RAsBAcH, K. & JEROME, C. (1995). Weitere Beobachtungen tiber das Vorkommen des Hautfarns Trichomanes speciosum Willd. in den Vogesen und dem benachbarten Deutschland. Carolinea 53: 21-32. RATCLIFFE, D. A. (1968). An ecological account of Atlantic bryophytes in the British isles. New phytologist 67: 365-439. RatcuiFFE, D. A., Birks, H. J. B. & Birks, H. H. (1993). The ecology and conservation of the Killarney fern Trichomanes speciosum Willd. in Britain and Ireland. Biological conservation 66: 231-247. Ray, J. (1724). Synopsis methodica stirpium Britannicarum, 3rd ed., pp. 119, 127-128. W. & J. Innys, London. REICHLING, L. & THorNn, R. (1997). Trichomanes speciosum Willd., un mystérieux passager clandestin. Adoxa 15/16: 1-4. Ricuarps, P. W. & Evans, G. B. (1972). Biological flora of the British Isles: Hymenophyllum. Journal of ecology 60: 245-268. Riptey, P. H. (1990). National field meetings, 1990: Brittany. British Pteridological Society bulletin 4: 11-14. Roserts, R. H. (1979). The Killarney fern, Trichomanes speciosum, in Wales. The fern gazette 12: 1-4. Rumsey, F. J. (1994). The distribution, ecology and population biology of Trichomanes speciosum Willd. Unpublished PhD. Thesis, University of Manchester. Rumsey, F. J., HEADLEY, A. D., FARRAR, D. R. & SHEFFIELD, E. (1991). The Killarney Fern (Trichomanes sodas in Yorkshire. Naturalist Jos 41-43. =anl DISTRIBUTION OF THE KILLARNEY FERN GAMETOPHYTE 19 Rumsey, F. J., HEADLEY, A. D. & SHEFFIELD, E. (1996). Gametophytic adaptations to low light — how unique are Trichomanes species? American journal of botany 83 (6) Suppl: 122. Rumsey, F. J., RAINE, C. A. & SHEFFIELD, E. (1992). The reproductive capability of independent Trichomanes gametophytes, in IbE, J. M., JerMy, A. C. & Paut, A. M., eds. Fern horticulture: past, present and future perspectives, pp. 299-304. Intercept, Andover. Rumsey, F. J., RAINE, C. A. & SHEFFIELD, E. (1993). Trichomanes venosum R.Br. (Hymenophyllaceae: Pteridophyta) in a Cornish garden — with a key to the filmy-ferns established in Britain and Ireland. Fern gazette 14: 155-160. Rumsey, F. J., RussELL, S. J., Ji, J., Barrett, J. A. & Gipsy, M. (1996). Genetic variation in the endangered filmy fern Trichomanes speciosum Willd., in Camus, J. M., Gipsy, M. & Jouns, R. J. eds. Pteridology in perspective, pp. 161-165. Royal Botanic Gardens, Kew. Rumsey, F. J. & SHEFFIELD, E. (1996). Intergenerational ecological niche separation and the ‘independent gametophyte’ phenomenon, in Camus, J. M., Gipsy, M. & Jouns, R. J., eds. Pteridology in perspective, pp. 563-570. Royal Botanic Gardens, Kew. Rumsey, F. J., SHEFFIELD, E. & FARRAR, D. R. (1990). British filmy fern gametophytes. Pteridologist 2: 40-42. SCANNELL, M. J. P. & SynnottT, D. M. (1987). Census Catalogue of the flora of Ireland, 2nd ed. Stationery Office, Dublin. SHEFFIELD, E. (1994). Alternation of generations in ferns: mechanisms and significance. Biological reviews 69: 331-343. Stokey, A. G. (1948). Reproductive structures of the gametophytes of Hymenophyllum and Trichomanes. Botanical gazette 109: 363-380. TEESDALE, R. (1800). Supplement to the Plantae Eboracenses. Transactions of the Linnean Society of London 5: IB), VOGEL, J. C., JESSEN, S., Gipsy, M., JERMy, A. C. & Eis, L. (1993). Gametophytes of Trichomanes speciosum (Hymenophyllaceae, Pteridophyta) in Central Europe. Fern gazette 14: 227-232. WIGGINTON, M. J. ed. (in press), The British red data book. Vascular plants. J.N.C.C., Peterborough. Yoroi, R. (1972). Studies on spore germination and gametophyte of Japanese Hymenophyllaceae. Science reports of the Tokyo kyoiku daigaku, Section B 15: 81-110. (Accepted June 1997) Watsonia 22: 21-28 (1998) 21 The use of B.S.B.I. Monitoring Scheme data to predict nationally scarce species in Britain D. A. PEARMAN The Old Rectory, Frome St Quintin, Dorchester, Dorset, DT2 OHF C. D. PRESTON, D. B. ROY I.T.E., Monks Wood, Abbots Ripton, Huntingdon, Cambs., PE17 2LS and A. STEWART D.E.R.C., The Barracks, Bridport Road, Dorchester, Dorset, DT1 1SN ABSTRACT It has recently been argued that the results of the B.S.B.I. Monitoring Scheme, a sample survey, can be used to identify species which may be nationally scarce in Britain. The practical usefulness of this method for estimating the frequency of uncommon species is discussed and shown to be limited by the large confidence limits associated with small sample sizes. The frequency estimates based on Monitoring Scheme data are tested for scarce species and for species in the Potamogetonaceae and Ruppiaceae, and are shown to under-estimate the distribution of many species. The revision of the list of nationally scarce species should await the results of the Atlas 2000 project, a geographically comprehensive project which will collate records collected over a longer time-span. KeEyworbs: plant surveys, mapping, rare species. INTRODUCTION In Britain, nationally rare species are currently defined as those present in 1-15 10-km squares (Perring & Farrell 1983) and nationally scarce species are defined as those in 16-100 10-km squares. We recently contributed to a review of the distribution of nationally scarce species (Stewart et al. 1994). The species selected for review were those on an existing list of scarce species (Nature Conservancy Council 1989), corrected and modified in the light of later information. This list is in turn based on the distribution maps in the Atlas of the British flora (Perring & Walters 1962), the last geographically comprehensive survey of the British vascular flora. In a recent paper, Rich (1997) has argued that the results of the B.S.B.I. Monitoring Scheme rather than those presented in the Atlas of the British flora should have been used to select the potentially scarce species. In the B.S.B.I. Monitoring Scheme, a sample of 1 in 9 of the British 10- km squares (or “‘hectads”’) were surveyed in two years’ fieldwork (1987-88). Rich (1997) dismisses as incorrect the suggestion of Stewart et al. (1994) that this survey was not designed to detect trends in relatively uncommon species, arguing that ‘“‘the sample survey as designed should have detected trends in all species, though clearly not as sensitively as a more detailed study, and less accurately for relatively uncommon species’’. He asserts that the results of the scheme ‘‘could have been related to Britain as a whole using standard statistical methods to provide a more up-to-date, rigorous selection of species to be investigated’”’. Rich suggests that as the Monitoring Scheme sampled approximately 1 in 9 10-km squares, the total number of squares in which a species occurs can be 22 D. A. PEARMAN, C. D. PRESTON, D. B. ROY AND A. STEWART estimated by multiplying the number of squares in which it was recorded in the Monitoring Scheme by approximately nine. To be precise, M x 2860 1 M M N 317 + 1.96 x 2860 x TC na ('- 35) where N is the number of 10-km squares containing the species expected nationally +95% confidence limits, M is the number of 10-km squares recorded in the Monitoring Scheme, 317 is the total number of 10-km squares covered by the Monitoring Scheme and 2860 is the number of 10-km squares in Britain. (The formula cited above is the one used by Rich, although it is cited incorrectly in his paper.) In this paper, we examine the limitations of Rich’s method as a predictive tool for rarer species. We then test the accuracy with which this method predicts the distribution of scarce plant species and of another group for which data have recently become available, the Potamogetonaceae and Ruppiaceae. RICH’S METHOD APPLIED TO UNCOMMON SPECIES Stewart et al. (1994) suggest that the results of the Monitoring Scheme should not be used to provide data on relatively uncommon species; Rich (1997) disagrees. The difference between the two points of view is almost certainly semantic rather than a real difference of opinion. Rich (1997) argues that the results will detect trends in the distribution of all species, but less accurately for the relatively uncommon species. We maintain that there comes a point where the accuracy is sufficiently low that one has to conclude that the method is not working, rather than working less accurately. There are, for example, 16 of the 253 nationally scarce species listed by Stewart et al. (1994) and an additional 145 out of some 310 rare species included in the Red data book (Perring & Farrell 1983) which were not recorded in any of the 10-km squares recorded for the B.S.B.I. Monitoring Scheme in 1987-88. It would be difficult to argue that the Monitoring Scheme has provided much useful data on these species. If a species is recorded in a few Monitoring Scheme squares, the prediction of the number of squares in which it occurs nationally is necessarily accompanied by large confidence limits. This severely limits the practical usefulness of Rich’s method for uncommon species. Even a species recorded in ten Monitoring Scheme 10-km squares is predicted as occurring in 90455 squares nationally, i.e. between 35 and 145 squares. This spans the range between species which are manifestly scarce to those which are much too frequent to qualify. Rich (1997) lists as potentially scarce all those species where the minimum prediction falls below 100 squares, as long as these are not known from other evidence to be more frequent. Only eight of the 65 species listed by Rich (1997) have a maximum predicted national distribution below 100 squares (Barbarea stricta, Callitriche brutia, Equisetum hyemale, Juncus ranarius, Monotropa hypopitys, Salicornia fragilis, Utricularia australis and U. ochroleuca). Six species (Calystegia soldanella, Eryngium maritimum, Glaucium flavum, Silene acaulis, Spartina anglica and Vaccinium uliginosum) which are listed as potentially scarce are estimated as occurring nationally in the range 96-246 squares. As Rich (1997) states, it is unlikely that many of the species will qualify as scarce. TESTS OF RICH’S METHOD SCARCE SPECIES We have compared the predicted distribution of the potentially scarce species included in Scarce plants in Britain (Stewart et al. 1994) with the known distribution of the species as reported in that book. The Monitoring Scheme results correctly predict the distribution of 231 species (72%) using Rich’s methods, over-estimate the distribution of one species (<1%) and under-estimate the distribution of 91 species (28%). There were 62 species considered by Stewart et al. (1994) which turned out to be present in more than 100 10-km squares, and were therefore too frequent to be considered as nationally scarce. The results of the Monitoring Scheme predict that 59 of these (95%) might be scarce, i.e. the minimum B.S.B.I. MONITORING SCHEME AND SCARCE SPECIES 23 prediction of these species is fewer than 100 10-km squares, and that twelve (19%) will be scarce, i.e. the maximum prediction for these species is fewer than 100 squares. (These 62 species are omitted from Rich’s paper as they were already known not to be scarce.) These results suggest a tendency of the Monitoring Scheme results to under-estimate the national distribution of species. POTAMOGETONACEAE AND RUPPIACEAE The accuracy of the predictions made by Rich’s (1997) method can be tested against another recently published dataset. In an account of the Potamogetonaceae and Ruppiaceae, Preston (1995) presented updated distributional data for the British taxa. The 24 species and the three commonest hybrids in these families are considered here. In compiling the distributional data, attempts were made to collect as many reliable records as possible for 18 species and the three hybrids, e.g. by contacting B.S.B.I. vice-county recorders. These are subsequently described as the well-recorded taxa (although they may not be well-recorded by the standards of other, more popular groups). The vice-county recorders were not contacted for records of six of the commoner species, and the 10-km square distribution of these species may therefore be underestimated. These six species are described here as the under-recorded taxa. The number of squares in which the well-recorded and under-recorded taxa have been recorded in the period from 1970 onwards is compared in Table 1 to the figures predicted from the Monitoring Scheme results. The Monitoring Scheme data are derived from Rich & Woodruff (1990) and the expected national totals calculated using the equation cited above. The results in Table 1 show that four of the 21 well-recorded taxa were not recorded at all in the Monitoring Scheme. The predicted number of squares falls below the recorded number for 15 of the remaining 17 taxa. For seven taxa, the recorded number is greater than the range predicted by the Monitoring Scheme results, based on the 95% confidence limits. If one assumes that there was no significant decrease in these taxa between 1970 and 1988, the results of this test also suggest that the Monitoring Scheme data tends to underestimate the national frequency of species. The Monitoring Scheme prediction exceeds the known 10-km square distribution for two of the under-recorded taxa, and it is almost certainly a more accurate estimate of their frequency. The estimate is below the recorded total for one under-recorded species, even though the recorded total is believed to be too low. POSSIBLE REASONS FOR UNDER-ESTIMATION USING RICH S METHOD The results discussed above suggest that the Monitoring Scheme results consistently underestimate the distribution of species. Rich (1997) implicitly assumes that the results of the Monitoring Scheme provide an adequate list of the species in the 10-km squares surveyed. There are two reasons to suggest that this assumption may not be justified: 1. There were two aspects to the botanical recording for the Monitoring Scheme. Recorders were asked to record the flora of three tetrads (2 x 2 km squares) within each 10-km square. They were also asked to record the species in the rest of the square. Our personal experience in recording for the Monitoring Scheme suggests that in areas where there were many botanists, both the specified tetrads and the rest of the squares were well recorded. However, in areas where there were few resident botanists, or which had to be recorded by visitors, the tetrads tended to be visited but the recording of the rest of the square was sometimes inadequate. This suggestion is supported by Rich & Woodruff’s (1990) analysis of the Monitoring Scheme database and by the data plotted in Fig. 1, which show that for a minority of 10-km squares almost all the records received came from the designated tetrads. The results of the Monitoring Scheme 10-km square survey are therefore likely to underestimate the number of 10-km squares in which a species occurs nationally. This does not preclude the use of tetrad rather than 10-km data to assess the national frequency of species. 2. Some species are likely to be under-recorded in a survey limited to two field seasons. These . include species which are difficult to detect in the field or to identify once found. Botanists with a particular interest in such difficult or critical species are much more likely to record them than those who do not have such specialised knowledge. The knowledge of such species is therefore likely to grow gradually as specialists in a county or country cover the area. In order to test whether species were under-recorded, we have examined the extent to which each nationally scarce species was D. A. PEARMAN, C. D. PRESTON, D. B. ROY AND A. STEWART 24 ‘uoMeur[dxe Joy 1x9} 9a¢ | OS FEV VE Op) ZL puis viddny 007 569 LST +68ET pST snyofiuosijod *g r8 TS bLL+Tbb 6b snqwyofsad “q £6 LS9 6CL +19 89 snqwuijzad * J eo 8LOL SSt+099T r81 SUDIDU * J Cw SZ CVL+1C8 16 sndsi42 U0JABOWDIOg {exe} popiosal-Japuy) 8E LV SC+81 ( psoyasa viddny (ee Lie Lele © Mz xX “¢ £8 £8 6V+CL 8 saployridy “dg ev 16 LI+¥6 I snyOflgyvs x “_ 0 cL 0 0 SNIUNA ‘J 88 61€ b6 +087 1€ snpisnd “g Iv OLT 6E+SP S snsuojavdd * cs St T9FLIT €1 snipofisnigo *g te IEE 8 SI+6 I snsopou *q SC 801 LELE € suaqiUu x “g @ LET 69+brl 91 SUIIN] * J WZ LEG ELFTIL ST SNAUIWDAB * J tele OTT Steseunte v said * LS 6 epFrs 9 smmsofuy “dq 0 I 0 0 snapkyida *g 0 O€ 0 0 snssa1quod * J 68 IZ 9F +9 i SNIVLO]OI “J 86 CCS ICL +v1S LS AACE CRAG cal LS 9EC LOFSET ST snudjp “q 0 IT = () 0 SNIOJ1jNIV UOJaSOWYIOg 06 IEILZG 8/7 +681 KG psuap DIPUY|UaOAy {exe pap1odas-|Ja A anjeA paarasqo jo (S66[ UoIss1g) SHU] IDUapyUod JUIdYIS satoads asejuasied se uonoipasd Ayyeuoneu % 66+ Ajjeuoneu sasenbs SULIONUOW Ul papsOdeI JUIIYIS BULIOINUO\ popsosas saienbs jo ‘oN jo Joquinu paiipelg saenbs Wy-Q] JO ‘ON AVEAOVIddNad UNV AVAOVNOLADOWVLOd AHL JO SUAANWAN YOU SANTIVA GFAUASTO HLIM V.LVd AWAHOS DNIMOLINOW WOU GILOIGdYd AONANOAWA TWNOLLVN JO NOSIYVdNOO ‘T FTAVL B.S.B.I. MONITORING SCHEME AND SCARCE SPECIES 25 900 800 700 400 300 No. of taxa elsewhere in square 200 100 0 100 200 300 400 500 600 No. of taxa in tetrads A, J, W FicureE 1. The total number of taxa recorded in tetrads A, J and W for each British 10-km square covered by the B.S.B.I. Monitoring Scheme, plotted against the number of additional taxa recorded during the Scheme elsewhere in that 10-km square. Squares which lack records from tetrads A, J or W were excluded from the analysis: the excluded squares are coastal squares without land in one or more of these tetrads or unrecorded inland squares. Number of species 71-80 % $1-90 % "94-100 % 0-10 % Sines 21-30% Wises 41-50 % "51-60 % 61-70 & Proportion of Monitoring Scheme squares in which species were recorded 1987-88 Ficure 2. The number of 10-km squares in which nationally scarce species were recorded during the B.S.B.I. Monitoring Scheme (1987-88), expressed as a percentage of the total number of Monitoring Scheme squares in which they were recorded between 1970 and 1995. The number of species (vertical axis) falling in successive 10% bands is plotted. 26 D. A. PEARMAN, C. D. PRESTON, D. B. ROY AND A. STEWART Number of species n Mt n ‘i He 1 tern : 11-20 % 24-30 % 31-40 % 41-50 % 51-60 % 61-70 % -80 81-90 % Proportion of Monitoring Scheme squares in which species were recorded 1987-88 Ficure 3. The number of 10-km squares in which Groenlandia and Potamogeton species were recorded during the B.S.B.I. Monitoring Scheme (1987-88), expressed as a percentage of the total number of Monitoring Scheme squares in which they were recorded between 1970 and 1995. The number of species (vertical axis) falling in successive 10% bands is plotted. recorded in 1987—88 in those 10-km squares covered by the Monitoring Scheme in which it is known (from Monitoring Scheme and other records) to have been present in the period 1970-1995. The results of this analysis (Fig. 2) show that there was a wide range in the efficiency with which Monitoring Scheme recorders detected scarce species. The same analysis for Groenlandia and Potamogeton species provides similar results (Fig. 3). It will be noted that many of the potentially scarce species listed by Rich (1997) are easily overlooked, difficult to identify or taxonomically critical, and are just the sort of species which are likely to be under-recorded in a “‘snapshot”’ survey or, in some cases, in any survey involving non- specialists. The species with a predicted number of 10-km squares below 100 include recently recognised segregates of Juncus bufonius (J. foliosus, J. ranarius) and Utricularia intermedia (U. ochroleuca, a plant not well understood even now by British botanists), species in the critical genera Callitriche (C. brutia) and Salicornia (S. dolichostachya, S. europaea, S. fragilis), rather inconspi- cuous plants such as Bromus lepidus and Epipactis purpurata, and species which show variation in flowering behaviour from year-to-year (Utricularia australis and U. vulgaris sens. strict.). Five of the eight species with a maximum predicted square total below 100 are included in this group. The case of the Utricularia species is particularly difficult, as flowering material can be identified easily but is rarely encountered; both species usually reproduce vegetatively. Even if a detailed survey reveals records from fewer than 100 squares, it is arguable that the species should not be regarded as scarce as there are post-1970 records of vegetative material from 242 10-km squares (Preston & Croft 1997), and these plants must be referable to one or other of the two segregates. TREATMENT OF DOUBTFULLY NATIVE SPECIES Rich (1997) describes as ‘“‘welcome and objective” the fact that Briza minor and Poa palustris are excluded as aliens from the list of scarce species. In this Stewart et al. (1994) followed Stace (1991). However, Rich (1997) lists Barbarea stricta as a potentially scarce species, as he himself (Rich 1987) regards it as probably native although Stace (1991) describes it as probably introduced. In the B.S.B.I. MONITORING SCHEME AND SCARCE SPECIES pi, controversial subject of native status there are numerous individual opinions, and the only practical course open to us as editors of Scarce plants in Britain was to follow a standard source. All botanists will, if given a chance, argue for the inclusion of some taxa and oppose the inclusion of others. Thus, one of us (D.A.P.) strongly favoured the inclusion of Briza minor as a native and another (C.D.P.) would have excluded Erodium moschatum as an alien, but we both agreed to set aside our personal opinions and follow Stace (1991). In Scarce plants in Britain our explanation for including Brassica oleracea as a scarce species is inadequate; we are therefore to blame for the fact that it has been misunderstood by Rich (1997). Stace (1991) regards this species as possibly native and we therefore include it as a scarce species. We accept that it is impossible in many instances to distinguish native from alien colonies, but as there are fewer than 100 10-km square records for all the established coastal colonies, native and alien, we list the species as scarce. The confusion arises as the author of the Brassica oleracea account (Richards 1994) considers that the species is introduced. Following our arguments would not result in the inclusion in the lists of rare or scarce species of taxa which are accepted as introductions in all their British localities. TREATMENT OF ARABLE WEEDS The particular and in some ways insuperable difficulties of dealing with arable weeds are discussed in Scarce plants in Britain (Stewart et al. 1994, p. 473). These difficulties have been highlighted by the publication, also in 1994, of a booklet by Wilson & Sotherton (1994). Whereas we published records of Ranunculus arvensis from 221 10-km squares from 1970 onwards and over 100 from 1980 onwards, Wilson & Sotherton believe that there may be as few as six viable populations left. Similarly they describe Scandix pecten-veneris as probably occurring now in fewer than 25 10-km squares, although there are records from 85 squares from 1980 onwards and 131 squares from 1970 onwards. In evaluating the status of these rapidly declining species we decided to alter the criterion for inclusion of scarce species and take records from 1980 onwards, rather than use the 1970 date employed for the other species. This decision was made on purely pragmatic grounds. We accept Rich’s (1997) view that it is inconsistent and subjective, but we believe that where circumstances differ, uniformity of treatment is not necessarily desirable. Retention of the normal 1970 cut-off date for these species, as Rich (1997) recommends, appears to us to be an unrealistic option. CONCLUSIONS We do not dissent from Rich’s (1997) view that the list of scarce species in Stewart et al. (1994) is provisional; all such lists inevitably are. We ourselves stated that there must be a strong possibility that the list would require revision in the light of current work on rare species (Stewart et al. 1994, p. 18) and that other formerly commoner species may now qualify as scarce (Stewart et al. 1994, p. 12). However, it would be a more efficient use of resources to await the results of the Atlas 2000 project (Pearman & Preston 1996), which will provide up-to-date geographically comprehensive data on the distribution of the British flora, rather than investigate the particular species listed by Rich in isolation. We also agree with Rich (1997) that it might be preferable to define rare and scarce species as a percentage of the British flora rather than in absolute terms, although this would require a co- ordinated look at both rare and scarce species. It will probably be desirable to assess the distribution of both rare and scarce species in a more small-scale unit than the 10-km square (Pearman 1997). ACKNOWLEDGMENTS We are grateful to Peter Rothery for help in the preparation of this paper, and to Arthur Chater for very useful comments on an earlier draft. 28 D. A. PEARMAN, C. D. PRESTON, D. B. ROY AND A. STEWART REFERENCES NaTURE CONSERVANCY Council (1989). Guidelines for selection of biological SSSIs. Nature Conservancy Council, Peterborough. PEARMAN, D. [A.] (1997). Presidential address, 1996. Towards a new definition of rare and scarce plants. Watsonia 21: 231-251. PEARMAN, D. A. & Preston, C. D. (1996). Atlas 2000 — a new atlas of flowering plants and ferns. British wildlife 7: 305-308. PERRING, F. H. & FARRELL, L. (1983). British red data books: 1. Vascular plants, 2nd ed. Royal Society for Nature Conservation, Lincoln. PERRING, F. H. & WALTERS, S. M. eds. (1962). Ailas of the British flora. Thomas Nelson & Sons, London. PRESTON, C. D. (1995). Pondweeds of Great Britain and Ireland. B.S.B.1. Handbook no. 8. Botanical Society of the British Isles, London. Preston, C. D. & Crort, J. M. (1997). Aquatic plants in Britain and Ireland. Harley Books, Colchester. Ric, T. C. G. (1987). The genus Barbarea R.Br. (Cruciferae) in Britain and Ireland. Watsonia 16: 389-396. Ricu, T. C. G. (1997). Scarce plants in Britain: have some been overlooked, and are others really scarce? Watsonia 21: 327-333. Ricu, T. C. G. & WooprurrF, E. R. (1990). The BSBI Monitoring Scheme 1987-1988. 2 volumes. Chief Scientist’s Directorate Report no. 1265. Nature Conservancy Council, Peterborough. RicHarps, A. J. (1994). Brassica oleracea L. var. oleracea, in STEWART, A., PEARMAN, D. A. & Preston, C. D., eds. Scarce plants in Britain, p. 61. Joint Nature Conservation Committee, Peterborough. Stace, C. A. (1991). New Flora of the British Isles. Cambridge University Press, Cambridge. STEWART, A., PEARMAN, D. A. & Preston, C. D., eds. (1994). Scarce plants in Britain. Joint Nature Conservation Committee, Peterborough. WILSON, P. & SOTHERTON, N. (1994). Field guide to rare arable flowers. Game Conservancy, Fordingbridge. (Accepted July 1997) Watsonia 22: 29-32 (1998) 29 Relationship between species richness and rarity in Welsh aquatic floras D. M. WILKINSON Biological and Earth Sciences, Liverpool John Moores University, Byrom Street, Liverpool, L3 3AF ABSTRACT Comparison of the biodiversity of sites, such as potential nature reserves, often has to use incomplete data. In such cases rules of thumb, such as a positive correlation between species richness and the presence of rarities, could allow more rational decisions to be made. Such a relationship is tested in two Welsh aquatic data sets. This analysis supports the general positive relationship between species richness and rarity; however in a lake data set no correlation was found between the species richness or number of rarities when the emergent and submersed plant communities were compared; 1.e. a site which was good for emergent plants was often not a good site for submersed plants and vice versa. It is suggested that the species richness/rarity rule only applies to tightly constrained, homogenous data sets. KEYworDs: nature reserves, biodiversity, botanical surveys, Pilularia globulifera. INTRODUCTION The preservation of biodiversity is a major aim of applied ecology (Wilson 1992). One of the approaches to this is to construct networks of nature reserves, though serious questions have been raised about this as a long term strategy in the light of continuing environmental change, both natural and anthropogenic (Hunter et al. 1988; Huntley 1994). Biodiversity is composed of a variety of measures such as species richness (number of species at a site), species abundance (population size of species of a site) and habitat diversity (Hambler & Speight 1995). In comparing sites, for example in selecting nature reserves, often the main data available are species lists for the better studied taxa (e.g. vascular plants, birds or butterflies). This raises a number of important questions. For example, does species richness of a well studied taxon such as vascular plants positively correlate with species richness of less well studied groups such as beetles (Wilkinson & Slater 1995)? Does species richness in a group correlate with the occurrence of rare species of that group at a site? Selecting sites on the basis of a high species richness could be a mistake if important rare species are found in species poor sites which would not receive protection were-such a criterion used (Hambler & Speight 1995). Such questions can only be successfully addressed in an empirical manner, by investigating the relationships between the occurrence of different taxa, species richness and rarity at a variety of sites. As a contribution to such a research agenda this paper examines the relationship between species richness and rarity in two aquatic flora data sets from Wales, U.K. THE DATA SETS Two published data sets on Welsh aquatic floras were used. . 1. A set of 36 shallow and often ephemeral upland ponds from mid Wales, ranging in altitude from 320-500 m (Slater et al. 1991). These ponds are of particular conservation interest as around half of them contain the aquatic fern Pilularia globulifera. This fern is endemic to Europe and threatened throughout most of its range due to loss of habitat through drainage. These mid Wales ponds are one of its strongholds and of international importance (Woods 1993). One site, Park Farm, which 30 D. M. WILKINSON TABLE 1. SPECIES FROM SEDDON (1972) LISTED AS IN NEED OF SPECIAL PROTECTION IN THE WELSH WATER AUTHORITY AREA BY PALMER & NEWBOLD (1983) Submersed and floating species Emergent species Apium inundatum Baldellia ranunculoides Ceratophyllum demersum Butomus umbellatus Elatine hexandra Carex acuta Isoetes echinospora C. acutiformis I. lacustris C. elata Lemna trisulca C. lasiocarpa Lobelia dortmanna C. riparia Luronium natans C. vesicaria Myriophyllum spicatum Cladium mariscus Potamogeton alpinus Eleocharis acicularis P. crispus Hippuris vulgaris P. gramineus Oenanthe fistulosa P. lucens Ranunculus lingua P. obtusifolius Schoenoplectus lacustris P. perfoliatus S. tabernaemontani P. pusillus Typha angustifolia Ranunculus circinatus Veronica anagallis-aquatica R. trichophyllus Sparganium angustifolium S. natans Subularia aquatica Eleogiton fluitans Callitriche hermaphroditica C. obtusangula Nomenclature follows Stace (1991). appears in the data set of Slater et al. (1991) has been omitted from this analysis as it is a lower, more species rich site in improved farmland and so is atypical of the rest of the data set. 2. A data set of plants from Welsh lakes including both upland and lowland sites (Seddon 1972). This data set is comprised of two subsets: i. submersed and floating aquatic plants from 70 Welsh lakes; and ii. emergent plants from 72 Welsh lakes. Rare species are those having low abundance and/or small ranges, and this raises the question of how low or how small? This is often affected by factors such as the area under study or the purpose for which the list is being constructed. A single objective definition of rarity which can be applied in all studies of all taxa is unrealistic. This is illustrated by the wide range of definitions of rarity which have been used in the past (reviewed by Gaston 1994). In this study a rare species is defined as any species listed as “in need of special protection in the Welsh Water Authority area” by Palmer & Newbold (1993). This is an inclusive list containing some quite widespread and locally common species which were however considered to be possibly declining and in need of protection. In the upland pond data set there were two species “in need of special protection’’, Pilularia globulifera and Apium inundatum. The lake data set contained a longer list of such species which are listed in Table 1. In the following analysis species richness refers to the total number of species at a site and rarity refers to the number of species listed as “in need of special protection” found at a given site. RESULTS AND DISCUSSION The correlations between species richness and rarity for each of the data sets are shown in Table 2. In each case there is a positive correlation significant at p<0-001, although the correlation coefficients range from 0-500 to 0-722. A number of previous studies on a variety of taxa have found rare species to be positively correlated with species richness (Gaston 1994). For example, Wheeler WELSH AQUATIC FLORAS a TABLE 2. RELATIONSHIP BETWEEN SPECIES RICHNESS AND RARITY IN WELSH AQUATIC FLORAS Total no. Total no. rare Data set Reference No. sites spp. spp. I Upland pools, Slater et al. (1991) 36 66 2 0-722 Mid Wales Welsh lakes Seddon (1972) 54 91 41 0-500 (all plants) Welsh lakes Seddon (1972) 70 4] 24 0-536 submersed plants Welsh lakes Seddon (1972) 2 50 il9/ 0-631 emergent plants Note: all correlations significant at p<0-001. (1988) found a highly significant (p<0-001) positive relationship between number of rare species and total species richness for British fen vegetation. In the present study the strength of the correlation varies between data sets. The highest correlation (r, = 0-722) is for the upland pond data set where there are only two rare species, Pilularia globulifera and Apium inundatum, which are significantly associated (77 = 13-38, p<0-01, n = 36) in the data set though it included some sites with Pilularia but no Apium and some with Apium but no Pilularia. In this case selecting species rich sites would be an effective way of selecting sites with rarities. The Welsh lake data set is of particular interest. Although there is a significant correlation between species richness and rarity for total species (i.e. all lakes where Seddon (1972) lists data on both submerged and emergent plants), if the submersed and emergent data subsets are compared (Tables 2 & 3 of Seddon 1972) then no significant correlations are found (correlation between submersed and emergent species richness r, = 0-144, not significant at p = 0-05, n = 54; correlation between number of rare submersed and number of rare emergent species r, = 0.029, not significant at p = 0-05, n = 54). Therefore good sites for emergent plants tend not to be good sites for submersed plants. This is true whether a “good site” is selected on the basis of species richness or by the presence of rare species. This suggests that correlations between species richness and rarity may be community specific, with a whole lake being too large a unit for analysis as it contains a number of very different communities (e.g. emergent and submersed). This requires investigation by further studies as it is important for comparing sites of nature conservation importance. It is of interest that for the upland pond data set Wilkinson & Slater (1995) found a positive correlation between plant species richness and a measure of general invertebrate species richness but no correlation between water beetle species richness and plant or general invertebrate species richness. This suggests that different taxa as well as different communities can show different patterns of species richness, so that a good site for water beetles may not be a good site for aquatic plants. These results suggest that caution is required in the make up of data sets when extrapolated from species richness data to rarity richness. For example the relationship between plant species richness and rarity in British fens identified by Wheeler (1988) means that selecting the most species rich sites would also select the sites with the most rarities. However if the exercise were repeated for all British mires then selecting by species richness would miss many rarities as most of the sites selected would be fens as these tend to be more species rich than ombrotrophic mires (Wheeler 1993), so rare ombrotrophic species would be missed. This underlines the suggestion in the data presented above that the species richness/rarity rule is only likely to work for tightly constrained, homogenous data sets. Such heuristic rules, tested on the relatively well known British flora, could be of great use in areas where the flora is less well known, such as tropical forests, which are thought to contain some 40% of the world’s flora, sometimes with over 300 species/ha’ (Archibold 1995). ACKNOWLEDGMENTS I thank Fred Slater and Ann Hemsley for collaboration in the Pilularia globulifera study and Alan Clapham and the referees for comments on drafts of this paper. 32 D. M. WILKINSON REFERENCES ARCHIBOLD, O. W. (1995). Ecology of world vegetation. Chapman & Hall, London. Gaston, K. J. (1994). Rarity. Chapman & Hall, London. HamMBLER, C. & SPEIGHT, M. R. (1995). Biodiversity conservation in Britain: science replacing tradition. British wildlife 6: 137-147. Hunter, M. L., Jacopson, G. L. & Wess, T. (1988). Paleoecology and the coarse-filter approach to maintaining biological diversity. Conservation biology 2: 375-385. Hunt ey, B. (1994). Plant species response to climatic change: implications for the conservation of European birds. [bis 137: 5127-5138. PALMER, M. & NEWBOLD, C. (1983). Wetland and riparian plants in Great Britain. Nature Conservancy Council, Huntingdon. SEDDON, B. (1972). Aquatic macrophytes as limnological indicators. Freshwater biology 2: 107-130. SLATER, F. M., HEMSLEY, A. & WILKINSON, D. M. (1991). A new sub-association of the Pilularietum globuliferae Tuxen 1955 in the upland pools in the mid-Wye catchment of central Wales. Vegetatio 96: 127— 136. STACE, C. (1991). New flora of the British Isles. Cambridge University Press, Cambridge. WHEELER, B. D. (1988). Species richness, species rarity and conservation evaluation of rich-fen vegetation in lowland England and Wales. Journal of applied ecology 25, 331-353. WHEELER, B.D. (1993). Botanical diversity of British mires. Biodiversity and conservation 2: 490-512. WILKINSON, D. M. & SLATER, F. M. (1995). Relationship between plant and invertebrate richness in upland ponds in mid Wales. The naturalist 120: 103-108. WILson, E. O. (1992). The diversity of life. Harvard University Press, Cambridge, Massachusetts. Woops, R. G. (1993). Flora of Radnorshire. National Museum of Wales, Cardiff. (Accepted March 1996) Watsonia 22: 33-39 (1998) 33 The current status of Rumex rupestris Le Gall (Polygonaceae) in England and Wales, and threats to its survival and genetic diversity R. E. DANIELS N.E.R.C. Institute of Terrestrial Ecology, Furzebrook Research Station, Wareham, Dorset, BH20 5AS E. J. MCDONNELL Dungeon Cottage, Cocklake, Wedmore, Somerset, BS28 4HB and A. F. RAYBOULD N.E.R.C. Institute of Terrestrial Ecology, Furzebrook Research Station, Wareham, Dorset, BH20 5AS ABSTRACT All U.K. sites from which Rumex rupestris Le Gall, a dock endemic to N.W. Europe, had previously been recorded, and which could be identified from herbarium specimens and Flora records, were surveyed. The objectives were to obtain up-to-date information on the distribution of this apparently declining species, the size of its extant populations and the suitability of previously occupied sites for re-introduction under the Species Recovery Programme. The amount of genetic variation, and its distribution within and between populations was measured using isozyme electrophoresis. Although a relatively large amount of the total variation found is held within individual populations, there is evidence of some isolation of populations by distance. Results also support the expectation that larger populations hold more genetic diversity. Possible threats to the survival of populations and to variation within the species are discussed. Keyworps: Shore Dock, distribution, isozymes, gene flow. INTRODUCTION Rumex rupestris Le Gall (Shore Dock) (Polygonaceae) is a rare and threatened European endemic growing in scattered populations along the coasts of Galicia, western France, the Channel Isles, south-west England and south Wales, with an outlying (possibly introduced) population on Anglesey. It grows above high water mark where there is a supply of fresh water. Some populations occur in wet crevices in the lower parts of cliffs, where fresh water discharges on to the upper parts of beaches or slumped head material. Others are found in seepage zones at the base of cliffs or in small pools on wave cut platforms. Less frequently, plants also grow along the margins of streams or ponds in dune slacks. Population size is usually small, frequently with fewer than ten individuals, growing as scattered individuals or small clumps. Throughout its range it appears to be vulnerable to © the same set of threats: habitat instability, human activity and small population size. The objective of the present study, sponsored by English Nature under the Species Recovery Programme, was to determine the current status of the species in Britain and the Channel Islands, assess the magnitude of the different risks and develop proposals for its reintroduction into some of its former sites. This paper reports the results of the first two of these activities. 34 R. E. DANIELS, E. J. MCDONNELL AND A. F. RAYBOULD METHODS Floras and herbaria were consulted in order to determine the locations of all past and present populations of R. rupestris in England and Wales, and the dates when last records were made. In late summer 1994 all known extant sites in Devon and Cornwall, based on data in King (1989) and information from local botanists, were visited and counts made of plants present (McDonnell 1995). Species associated with colonies of R. rupestris were also recorded. Similar counts were made of populations in the Isles of Scilly by R. Parslow (Parslow & Colston 1994). Q. Kay and A. Jones provided information on the present status of the species in Wales. In summer 1995 most of the mainland sites were revisited together with all sites in Cornwall, Devon and Dorset for which old records had been found in order to determine whether any plants were still present (Daniels, McDonnell & Moy 1996). Comparisons were also made with sites visited the previous summer, in order to assess suitability for re-introduction, bearing in mind the known or predicted ecological requirements of the species and practical considerations. Such information, together with site history, where this was known, was also used to suggest possible reasons for loss of individual populations and to select sites for re-introduction. In 1994, small seed samples were collected from all populations visited. A sub-sample of these seeds was used to grow plants for examination of electrophoretic variation in leaf material. Enzyme extracts were prepared in 0-1M Tris buffer containing 10% glycerol, 1% ascorbic acid and 0-1% mercaptoethanol. The natural acidity of the leaf tissue obviated the need for pH correction of the buffer to pH 7-0 using HCl. The resulting slurry was centrifuged at 1400 rpm for 3 minutes and the supernatant was stored at —73° C until used for electrophoresis. Electrophoresis was performed in a BioRad Protean II chamber using 1 mm thick polyacrylamide gels. The 13% separating gel was prepared in a 0-4M Tris-HCl buffer (pH 8-8) and a 7% stacking gel in a 0-1 Tris-HCl buffer (pH 6:8). Gels were run for 7 hours at a constant 250V using a 0-072M glycine-0-005MTris electrode buffer (pH 8-5). Gels were stained for eight different enzyme systems using the protocols of Raybould et al. (1991). RESULTS STATUS AND DISTRIBUTION The distribution of extant (in 1989-1995) and former sites in England and Wales is shown in Fig. 1. Subsequent to this survey, a new site was found in Glamorgan, south Wales (Kay 1996). Table 1 gives the number of plants present in each of the surveyed sites in Devon and Cornwall in 1989 (data from King 1989), 1994 and 1995, together with 1994 and 1996 data for the Isles of Scilly (Parslow & Colston 1994; Parslow 1996) and 1996 data from Glamorgan (Kay 1996). Some difficulty was experienced in assessing absolute numbers because of possible confusion of vegetative plants with non-fruiting specimens of Rumex crispus, especially in its coastal variant, subsp. littoreus (J. Hardy) Akeroyd. Because of this, reliable comparisons of population size can only be made using counts of fruiting plants. Sites at which the plants had been recorded formerly but were not found in 1994 are listed in Table 2, together with the dates of last records. In some cases the actual locations are ill-defined because of lack of clarity in recording the precise location of a Flora entry or a herbarium specimen. In a few cases it has been assumed that two records with different names have referred to the same populations, especially where one of them has been as vague as “near Plymouth”. Whilst some locations appear several times, with the species being noted by more than one recorder, others are represented by single records only. GENETIC VARIATION Pot-grown plants at Furzebrook showed a wide range of variation in height, differences in leaf characters and divergence in inflorescence structure. In particular, nine robust plants with leaves standing out widely from the stems and wavy leaf margins (five from seed collected at Rame and four from seed collected at Westcombe) developed tall flowering shoots in which the branches formed acute angles with the main axes and the flowers were arranged in closely-set whorls. In _ addition, the fertile fruits produced had distinct wings. These plants so closely resembled R. crispus STATUS OF RUMEX RUPESTRIS 35 Ficure 1. Distribution of extant (closed circles) and historical (open circles) populations of Rumex rupestris in mainland Britain, Anglesey and the Isles of Scilly. 36 R. E. DANIELS, E. J. MCDONNELL AND A. F. RAYBOULD TABLE 1. NUMBER OF PLANTS OF RUMEX RUPESTRIS (IN FRUIT AND IN TOTAL) AT DIFFERENT LOCATIONS IN ENGLAND IN 1989, 1994, 1995 AND 1996 A zero indicates that no plants were observed: a dash indicates that no observations were made. 1989 1994 1995 1996 Site + fruit total + fruit total + fruit total + fruit total Cornwall Penhale - >50 33 35 _ >50 >70 >70 Lamorna - = 2 3 2 3 _ = Gunwalloe - >50 7 mn 0 p} - = Pendower ~ 18 9 9 9 9 - = Pencarrow - - 3 S38 Ti qf - = Llantivet Bay 1 - 2 2 ~ 1 +. - ~ Llantivet Bay 2 — — - - iT 7 - - Rame ~ 69 24 36 31 35 = = Devon Wembury - - = ~ 1 1 1 1 Westcombe 2 4 2 2 = - = = Soar Mill 5 6 13 15 3 8 - = Rickham - 19 0 0 0 0 - = Venericks ~ 9 5 13 10 12 _ = Isle of Scilly Tean _ - 19 i. - = 19 97 Tresco - - 33 3% - - 32 60 Samson - - 40 40 - ~ 90 124 Annet - - 6 6 - ~ 51 51 St Agnes - - 14 14 ohn - 18 18 Wales Newborough - ~ —. C25 ~ - _ - Glamorgan — — _ - - ~ 21 >21 that it appears most likely that in the field specimens of R. crispus subsp. littoreus were sampled rather than R. crispus, despite conviction at the time that all seed came from shore dock plants. A second group of plants grown from seed collected at Church Cove had spreading, leafy, inflorescences and comparatively long petioles and were, undoubtedly, R. conglomeratus Murray. Field survey in 1995 confirmed that all plants at this site were R. conglomeratus. Several individuals showed characteristics which were intermediate between R. conglomeratus plants and more typical R. rupestris in the cultivation trial. This suggests that either the range of variation is continuous between the two species or that some of the plants were of hybrid origin. Of the eight enzyme systems tested, five produced only monomorphic banding patterns and only three (esterase, malate dehydrogenase and phosphogluco-isomerase) showed polymorphism. Even in these cases, band interpretation in terms of alleles at different loci was difficult and different band combinations were recognised only as different phenotypes. Because of this, variation at the species and population levels was calculated using the Shannon Diversity Index (King & Schaal 1989) rather than F-statistics. Each of the phenotypes for each of the systems where polymorphism was found were combined to give a set of overall phenotypes (e.g. AAA represents a phenotype showing banding patterns recognised as phenotype A for each of the systems whilst AAB represents a combined phenotype with a different PGI banding pattern). The proportions of each of these combined phenotypes were used to calculate dissimilarity indices. Calculations of overall diversity (H,,), the mean diversity within populations (mean of H,.,) and the contributions of within-population and among- population to overall diversity were made at three different scales. Taking all populations as distinct, independent, units, 55% of the total variation was contained within populations and 45% was attributable to differences among populations. This implies that, although there are differences among the isolated populations of R. rupestris in south-west England, each still retains much of the total variability found within the species in England. The maintenance of diversity may result from gene flow between populations (either through pollen transfer or seed dispersal) and if this is the STATUS OF RUMEX RUPESTRIS 37 TABLE 2. FORMER SITES OF RUMEX RUPESTRIS INENGLAND AND WALES English sites with asterisk are those where there is only a single record. Site Grid square Last record Cornwall Harlyn Bay SW/8.7 1900* (Trevose Head & SW/8.7 1963) (Constantine Bay 1951) East Pentire SW/7.6 Fistral SW/7.6 1912 Newquay SW/8.6 Gravel Hill mine, Cubert SW/7.5 1903* Godrevy Point SW/5.4 1951* Lelant SW/5.3 1909* Sennen Green SW/3.2 1870’s* Boscawen. Cliff SW/4.2 1900* Poltesco - SW/7.1 1870’s Hemmick Beach SW/9.4 1905* Vault Beach SX/0.4 1900* Looe SX/2.5 191," Downderry SX/3.5 1875* Portwrinkle SX/3.5 1875* Tregantle SX/3.5 1875* Devon Wadham SX/5.4 1875* Pamflete SX/4.5 1876* Little Seacombe SX/7.3 1989 Gammon Head SX/7.3 before 1939 Slapton Ley SX/8.4 1977* Braunton Burrows SS/4.3 1955 Dorset Lyme Regis S/3-9 1923* West Bay SY/4.9 1949* Ringstead Bay SY/7.8 1985 Durdle Door SY/8.8 1985 Poole SZ/0.9 1900* Glamorgan Three-cliffs Bay SS/5.8 1985 Pennard Burrows SS/5.8 1910 Kenfig Burrows SS/7.8 1948 Dunraven Bay SS/8.7 1934 Methyr Mawr SS/8.7 1954 Pembrokeshire Lydstep Haven SS/0.9 1957 case we would expect the highest rates of gene flow to be between neighbouring populations and the lowest between the most distant populations (isolation by distance). Gene flow could not be implied directly from our results because of our inability to define alleles and our resulting reliance on calculation of dissimilarity from phenotypic data only. Pairwise calculations were made to determine the relative amount of variation found within pooled pairs and between them. The results were then compared with geographical distance (log transformed to allow for the wide range of distances used) between the population pairs. A significant positive relationship was found between geographical distance and the amount of dissimilarity shown by the population pairs (b = —2-98 x 10-; p (b>0) = 0-035). The further apart the populations were, the more dissimilar they were. Separating Scilly Isles populations from those on the mainland and performing partial regressions on the results to allow for the fact that populations within each of the two regions are closer to each other than populations in different regions showed highly significant distance and region effects. In other words, the effect of distance between populations (after removing effects due to their 38 R. E. DANIELS, E. J. MCDONNELL AND A. F. RAYBOULD presence in different regions) on their dissimilarity was highly significant (b = 5-77 x 1077; p (b>0) = 0-0014). Dissimilarity between regions (Scilly Isles and mainland), once distance effects had been removed, was also significant (b = 4-41 x 107°; p (b>0) = 0-0056). Although individual populations do hold much of the variation, we might expect that the amount would be reduced as population size decreases so that dispersed, small, populations would be susceptible to gradual loss of genetic variation. A regression of diversity on population size did show a positive relationship (r7 = 34-3%) so confirming this expectation. DISCUSSION The fact that the list of extant sites is shorter than that for former sites suggests a sharp decline in distribution of the species. In some cases the reasons for loss of Rumex rupestris from particular sites are quite clear. Drastic habitat modification caused by sea wall construction and cliff consolidation are known to have been instrumental in the decline of R. rupestris from sites such as Gunwalloe and may have been major factors in eliminating the plants from sites such as Lelant, Hemmick Beach and Slapton. Elsewhere, erosion, especially under the influence of severe storms, may have been responsible for colony destruction, as appears to have been the case at Ringstead Bay. Plant community development (especially the growth of a closed sward of coarse grasses and herbs) may also have been a contributory factor to the loss of R. rupestris from sites such as that at Constantine Bay. The occasional occurrence of a single plant, or a few, isolated, individuals in sites from which they subsequently disappear may indicate that many populations are inherently ephemeral. Conditions for successful germination and seedling establishment may occur only occasionally, with those plants which do become established surviving for one or several years depending on site stability. At the same time, the appearance of plants in new locations from time-to-time does suggest that either seed banking and subsequent exposure, or seed dispersal, is effective and that seeds do occasionally arrive at suitable germination sites. Such a dynamic situation is not uncommon among species (including Rumex spp.) which do bank seed and require re-exposure before they can germinate successfully (Kendrick & Heeringa 1986; Voesenek et al. 1992). However, the unstable coastal habitats in which R. rupestris grows do not appear conducive to the retention of a persistent seed bank and so the potential for recolonisation will depend on a supply of seed from neighbouring populations reaching suitable germination and establishment sites. Given suitable sites we know from practical experience that seedling establishment is possible. However, as individual popu- lations are extirpated or as conditions lead to the progressive loss of suitable germination or establishment sites, the capacity for replenishment is decreased. In a species which, with a few exceptions, appears to occur predominantly as small, isolated, populations, there are clearly risks associated with that small population size. Individuals become relatively more important as significant components of the population and the risks of loss of genetic diversity are increased. Isolation will also limit gene flow between individuals in different populations, unless there are effective mechanisms for pollen or seed dispersal, and so reduce the capacity to replace lost variation. The existence of a high level of genetic diversity within populations and within groups of populations suggests that isolation may not be complete and that each population is a random selection of genes present in a wider gene pool. As the distance between populations, or groups of populations, increases, the possibilities for gene exchange are decreased and local patterns of variation become established. This is shown in particular when the Scilly Isles populations are compared with those on the mainland. When introducing or reintroducing the species to locations it may be more effective to use seed from more than one population in the region of the new site in order to obtain a more complete sample of the variation present in that region. Seed collected from a single (especially a small, single) population may represent only a small part of the variation available in the vicinity of the establishment site. These results are only preliminary and more work is required: firstly, to develop systems for measuring genetic variation which can be interpreted more accurately in terms of allelic variation; secondly, to determine differences between plants in south-west England and other parts of the species’ range; and thirdly, to ascertain whether the high levels of variation found at a local scale are STATUS OF RUMEX RUPESTRIS 39 related to gene flow between individual populations at a local scale or to introgression of genes, for example, from the closely related R. conglomeratus. ACKNOWLEDGMENTS This work was carried out as part of English Nature’s Species Recovery Programme and we are grateful for the financial assistance provided. We would also like to thank Miles King (Plantlife) and Roger Mitchell (English Nature) for their help, especially in the early stages of the project, and Simon Creer for carrying out the electrophoresis. REFERENCES DanieELs, R. E., McDONNELL, ©. J. & Moy, I. L. (1996). Species Recovery Programme: shore dock (Rumex rupestris Le Gall). Second report. Report to English Nature, Species Recovery Programme. Kay, Q. O. N. (1996). The conservation of Rumex rupestris (Shore Dock) in Wales. Past, present and possible future sites and habitats of Rumex rupestris in South and West Wales. Report to Countryside Council of Wales. Kenprick, R. E. & HEERINGA, G. H. (1986). Photosensitivity of Rumex obtusifolius seeds for stimulation of germination: influence of light and temperature. Physiologia plantarum 67: 275-278. Kina, L. M. & ScHaat, B. A. (1989). Ribosomal-DNA variation and distribution in Rudbeckia missouriensis. Evolution 43: 1117-1119. Kina, M. P. (1989). An investigation into the current status and ecology of the shore dock Rumex rupestris in Devon and Cornwall. M.Sc. thesis, University College, London. McDonnell, E. J. (1995). The status of shore dock (Rumex rupestris Le Gall) in Britain in 1994. Report to English Nature, Species Recovery Programme. ParsLow, R. (1996). Species Recovery Programme: Shore dock Rumex rupestris Le Gall in the Isles of Scilly. Report to English Nature. ParsLow, R.& Corston, A. (1994). The current status of Rumex rupestris in the Isles of Scilly. Report to English Nature, Species Recovery Programme. RAYBOULD, A. F., GRAY, A. J., LAWRENCE, M. J. & MARSHALL, D. F. (1991). The evolution of Spartina anglica C. E. Hubbard (Gramineae): origin and genetic variability. Biological journal of the Linnean Society 43: 111-126. VoESENEK, L. A. C. J., DE GRAAF, M. C. C. & Bom, W. P. M. (1992). Germination and emergence of Rumex in river flood-plains. II. The role of perianth, temperature, light and hypoxia. Acta botanica Neerlandica 41: 331-343. (Accepted May 1997) oct ta: CMscdennibericty i Sr wake Hegibus (3 ie SEMEN edt. ah me SET i ‘eat cb - & eae Ta aT ern Pint i Laird wey ly a RB PA tie ae: aaa an wie bos (otiltr oe nk ; @) sali) oels DbOW ® ote % bare nejorg aft 16 — vhse’ suf ai lista ats it seit F oe er eee es ae imeas - = 4 a) ec: a Be en Hn ant: ie ea avis fe Ris Weise . 2 38 Tessas » Stig natin ti: eer es ilk Mh! t 5 ee : zee Ce aan’ on re 4 | bP 5. 1 morn x0 ro aude Akay cessed jessie . . | STEER acral ipdieireiis Peck ie wie ethwA ai nonyditead has dotetay AAC ssiomiee 6 ee ood Alt < : Veo ge ed eae ree 4 ysis ALES ed a spears. asp ode won {3 % cagten ee) aerate / be wi! FN eee Wet aca Naaree ii sy is Fy Sot eh ‘ee OTN i rena t 2) cagitiohy What fads ei y on LAR ry 1 M, noah 4 a pe ied wetted. Fa se bey ar Miao or alcelarsieas ah 7 ant | Th VAG ae Seay, Amie tet ate) BOE on ie 7 ye ve +7) ¢ th aA a Ons alee 4 \ heen ' no ytte Siow between ae ass uti pines 8 moh) oF oc eee Mm or uy Bae’ | et ee ie a cont janet ome treat ths emi et Ee AN Bs nth HS : race: aap my ee eee ee iy . a SOR ee ty Ad oH . re Dia asih pre . “by a SS ae OTE age eT aaa Sia in : # fikdtet tet / 20 beg ; habia tt Larrea Peep Me =9¢ Se ae eee chee ea Mt ee De act OW bs ied ibe. st e ies |. ehewte } WORE Bie o Pone x avid # smal § ie ate Ea oe 1 Epes, ‘ Rae ee HX oF Peper, hive, we ag: wa year’ brs ewig eis Vee, ETE ee’ CRE as SEP wiv vig a ota ay « ecnala vos end thiedl/ oakwereh ehbihes Gee tie teh ie fone acein ‘ania an Ae = » 2 ~ < a sD ts Watsonia 22: 41-47 (1998) 4] The distribution and status of Corynephorus canescens (L.) P. Beauv. (Poaceae) in Britain and the Channel Islands with particular reference to its conservation P. J. O. TRIST Glovers, 28 High St, Balsham, Cambridge, CBI 6DJ ABSTRACT Corynephorus canescens (L.) P. Beauv. is considered to be a scarce plant in Britain. A description of the plant and its habitat are outlined with a brief account of each site. The local problems in the maintenance of the population at the inland site at Wangford in the W. Suffolk Breckland (v.c. 26) are discussed with accounts of conservation treatments. KEYWORDS: conservation, sand accretion, Carex arenaria, pH range, vegetative spread. INTRODUCTION Corynephorus canescens (L.) P. Beauv. (Poaceae) is considered to be a scarce plant (i.e. it occurs in 16-100 10-km squares) in Britain (Trist 1994). This paper is concerned with providing a description of the species and an up-date on its distribution including comments on its habitat, population maintenance and conservation measures. DESCRIPTION A densely tufted glaucous perennial of variable life span, 5-30 cm high with fibrous roots. Culms mostly erect, 2-5 noded below the middle. Leaf sheaths inflated, light purple; ligules acute, 2-4 mm; blades numerous, glabrous, stiff, setaceous, inrolled, (1-5)-4-12 cm x 0-3-0:5 mm. Panicles narrow, open at flowering, later becoming dense, 0-8-7 cm X 0-5—1-5 cm, pale green or purple; branches short; pedicels 1-3 mm long. Spikelets lanceolate, compressed, 2-flowered, 2-5-4 mm long, breaking up at maturity below the lemma. Glumes narrow, lanceolate, acuminate, equal or almost so, light purple, membranous, slightly scabrid on the keel, with narrow white tips and margins. Lemmas enclosed by the glumes, 1-5—2 mm long, ovate, with a basal tuft of minute hairs and awned at the base. The awn has two parts, in the middle is a ring of hairs, the lower half is orange — brown and twisted when dry, the apex is club-shaped and is mainly enclosed by the glumes but 0-5—0-8 mm of the length is visible. Paleas about as long as the lemmas. Anthers 1-0—1-5 mm long, mostly deep violet. Chromosome no. 2n = 14. FIELD RECOGNITION On close approach, a C. canescens colony is seen to consist of small single tufted plants with no grouping or display of a sward. In early July plants can be seen 30 to 40 m distant when the panicles are spreading at anthesis and the colony is a mass of silver and pale pink colour. The stiff pointed leaf blades which form a dense cluster are a bright metallic green with the sheaths a pink-purple colour. In the autumn the tufts die back to a grey colour and the young green leaves grow from a bud 42 P. J. O. TRIST low on the culm which is hidden under the growing plant. A diagnostic feature is the shape and division of the awn. HABITAT Nomenclature for vascular plants follows Stace (1997) and for mosses, Smith (1978). C. canescens is found on tidal beaches, on fixed and mobile dunes and sometimes on coastal sandy turf: it is also known on isolated inland sandy heaths and dunes. It is essentially a plant of open conditions where wind can move and promote sand accretion around the plants. An internode low on the stem responds with vegetative elongation following partial covering by sand. It does not spread vegetatively and the new elongation only forms a replacement plant. Individual clumps of plants in the Breckland sites are rare outside the immediate area of a colony, as few plants are derived from seed. Its position on the coastal sands is often within 100 m of the tide and well within reach of salt spray and we may assume it has a fair salt tolerance. Marshall (1967) reported that the habitats range from base poor soils to calcareous dunes within a pH range of 4-0-8-3 and also recorded that it is a plant of substrata which are extremely low in mineral nutrients. At Hedderwick, E. Lothian, v.c. 82, numerous samples were taken at ten stations at a depth of 6-105 cm on beach sand with a high proportion of shells. The pH ranged from 6-0-8:7, with one exceptional pH 9-08 and an outlier off the shells of pH 5-18. At Toigal, Westerness, v.c. 97, an equally high pH 8-3 was taken from a shell dune, which also demonstrated very low mineral nutrients with available phosphorus at 4 mg/l, potassium 9 mg/l and magnesium 7 mg/l. At Wangford, W. Suffolk, v.c. 26, the dunes are of coarse brown very acid sand, with a pH 4-6. The sand is mainly of quartz, stained with iron with a very small fraction of hematite black grains and no organic matter. In various areas Corynephorus canescens is accompanied by a number of different taxa but the number of constant species per site is generally small. Carex arenaria is present at most sites and Ammophila arenaria is generally present at maritime sites. CONSERVATION The study site at Wangford Glebe is 2-5 km from Brandon in the W. Suffolk Breckland and is an S.S.S.I. owned by the Suffolk Wildlife Trust. Corynephorus canescens occupies a small fixed dune of 7 X 9m which in 1995 was c. 80 per cent bare sand and invaded by Carex arenaria on all sides. Agrostis vinealis is its main constant species with a little Polytrichum piliferum occurring with Cladonia spp. It has recently been noticed that the annual disturbance caused by sand spreading is having the effect of reducing the Cladonia. This site was found by M. G. Rutterford and myself in January 1970 (Trist 1971) and at that time it was thought to be extinct in Breckland. This find was then the only known inland site of Corynephorus canescens in Britain. Hind (1889) notes that Druce and Bolton King recorded C. canescens in July 1883 “‘between Lakenheath and Wangtord”’. Druce also collected specimens from the same site in August 1883 (OXF). This site clearly refers to the part of Lakenheath Warren now occupied as Lakenheath Airfield, where several large colonies of C. canescens survive. Wright (1668) recorded severe sand storms in Breckland, which may have caused the large blow- out on Wangford Glebe. Sand storms still occur on arable land in Breckland and can do severe damage to recent seeding and pile up sizeable ridges of sand, but the storms are probably not so fierce as in the past. The taming of the wind has considerable significance in relation to the present conservation management of C. canescens. Some 70 years ago, the area from Lakenheath to Brandon was a vast open sandy heath with few trees and subject to frequent wind erosion. Large conifer plantings by the Forestry Commission in the 1920s have grown into mature trees and halted the distances over which erosion takes place, although it can still occur on the Wangford dunes which represent the last of the sites of active erosion in Breckland. This increasing wind protection has halted sand erosion around the Wangford colony of C. canescens. So active erosion is being effected by the Trust with annual rotovations in certain areas of the reserve. This colony is above a line of underground spring water which runs from a known site in the forest on the east of the reserve and follows its northern boundary to discharge eventually into the Wangford drain. This latter is over 3 m deep and in July 1992 was dry. The source of this water is probably DISTRIBUTION AND STATUS OF CORYNEPHORUS CANESCENS 43 being diminished by a deep bore pump within 1-5 km of this site. As proof of its line, there were two very small colonies of Juncus squarrosus and a dozen or more plants of Juncus effusus which thrived in a water fed zone over an iron pan within the blow-out area and near the northern boundary of the reserve. An area some 250 m away in the centre of the reserve had up to 15 old tussocks of Carex paniculata and a bed of Calamagrostis canescens; these plants were no longer found after c. 1984. This loss of moisture is an additional factor in the conservation of the Corynephorus canescens site as it is encouraging the invasion of Pteridium aquilinum, Deschampsia flexuosa and Carex arenaria, all of which are increasing in the area of the C. canescens. The one-time open habitat requires management or it will be invaded to the exclusion of C. canescens. When the colony of C. canescens was first found in 1970, 231 plants were recorded. No detailed recording took place in the following years, but observations between 1972 and 1974 showed the population had apparently declined to about 150 plants. In the autumn of 1975 a count revealed that the population was reduced to 112 plants. It had been a dry summer and a few seedling plants were noted. A few of the dried up plants were showing young vegetative growth (Trist 1980). At this time, there was no sand movement by wind in the area of the C. canescens colony. Marshall (1967) had noted that C. canescens is found in disturbed or open sandy places where up to 10 cm per year of accretion takes place: and while the adult plant cannot withstand complete burial, it responds to partial burial by vegetative elongation. Accordingly, in the autumn of 1975, artificial sand accretion was introduced by cutting a wide trench at the foot of the Corynephorus canescens sand bank to aid wind disturbance of sand. In addition, a good covering of sand was spread over the colony. This was repeated in the autumn of each year up to 1978. In Trist (1980) the report showed a count of 422 plants which I now consider inaccurate. The 232 yearling plants would be accurate but, “‘a further 190 flowering tufts, many of which were developing new vegetative shoots’ should be disregarded. The total count should have been assessed at about 250 plants. The management of the colony has both changed and lapsed. Between 1978 and 1992 no sand was applied to the plants and no counts were made until 6 February 1992 when 226 plants were recorded. The area was sanded on 7 March 1992. The next count was made on 17 June 1993 and showed 254 plants. Sand was again applied on 19 November 1993 and a count taken on 7 September 1994 revealed 265 plants. Later I did not feel satisfied that this was an accurate record, as at the time I recalled there were some plants in the process of dying back which were not showing the tuft of new leaves at the stem node. On 3 November 1994, a further count gave a total of 304 plants. The summer of 1995 was hot and dry and on 29 November 1995, a count of 142 plants also proved an inaccurate record, as there was an unusual number of seedling plants of C. canescens and other grasses which could not be separated at the young stage. Recording in September 1996 gave 303 mainly small plants with only one seed head. Rabbits may play a small part in the conservation of C. canescens. They will eat young shoots of Carex arenaria but as long as other food is available, their attention to C. canescens is confined to biting off a few inflorescences which matters little as the loss of potential seed is probably of little account. At Wangford, propagation of the species is mainly by vegetative elongation. Since this colony was found in 1970, we have always had rabbits in residence. In 1995 there were two active burrows, the occupants of which contributed to sand disturbance. The conservation of C. canescens is complicated. Some problems may be resolved while the solutions to others are impracticable and some not possible. The inland sites are subject to changes of their immediate surroundings by agriculture and other land uses. I agree with Marshall’s (1967) conclusion that ‘“‘many of the present day European communities of Corynephorus cancescens owe their existence to human interference’. This comment aptly applies to past and present events at Hedderwick, Lossiemouth, Kessingland and Lakenheath, which are recorded below. The coastal sites are at all times open to the threat of gales and tidal erosion. SITE DESCRIPTIONS THE COASTAL SITES Suffolk has three coastal sites. These are considerably smaller in area and in population than those of Norfolk. At Minsmere Haven the area of 40 X 3 m is a restored site c. 230 m from the foot of 44 Pi. G.On TRISTE Minsmere cliffs. The former site on dunes near the base of the cliffs was destroyed in the sea floods of 1953. This new colony comprises 250 plants on the landward side of a sand bank between the boundary ditch of the R.S.P.B. reserve and the beach shingle ridge. The site is still within danger of a high tide. In 1992 the colony had no constant species. At Kessingland our plant is sparsely spread out over beach sand and has to tolerate the summer activity of an adjacent holiday camp. Catapodium marinum, Phleum arenarium and Vulpia bromoides are scattered associates. The site on the beach near Benacre Broad is spread over c. 0-16‘ha. This colony flourishes in spite of considerable tidal approach as the land by the east coast inclines. In the past 30 years some gale force tides have crossed the beach to the land. The Norfolk coastal sites are extensive, on tidal beaches and on hinterland dunes. Their total population of C. canescens would number many thousands. From Great Yarmouth dunes where this grass is abundant, north up the coast to Caister-on-Sea, the line of plants is almost continuous on the beaches for nearly 3 km. Just beyond the Caister lifeboat station, C. canescens colonises narrower dunes within the Caister golf course for a distance of c. 200 m. Our plants again occur to the south of Winterton-on-Sea where they also colonise hinterland dunes in comparative shelter where Ammophila arenaria and Carex arenaria are abundant. It is here that the dispersal of C. canescens is noticeable and which appears to indicate moisture competition with the above two species. From Winterton our plants continue north-east to Horsey Common and then abruptly stop. They reappear at Blakeney where there is a scattered population from west to east over to Blakeney Point, a distance of c. 800 m where our plant is frequent on grey dunes. There was a colony on the beach at The Hood, Blakeney which has a long and difficult approach over shingle. There are no recent records. The constant species of the various sites north of Great Yarmouth up to Blakeney are fairly similar. Jasione montana, Hypochaeris radicata and Rumex acetosella are frequent and often with Festuca arenaria. Beyond the Caister lifeboat station the dunes are smaller and Jasione montana is replaced by Campanula rotundifolia. At Holme-next-the-Sea our plant occupies a small open dune on a slope surrounded by Ammophila arenaria. Some Corynephorus canescens is seen at the base of the slope within reach of high tides as indicated by Glaux maritima and Silene uniflora. The colony has a population of c. 300—400 plants. All of the Norfolk sites are much exposed to wind erosion and several stretches of Corynephorus canescens between Great Yarmouth and Caister are occasionally subject to tidal wash. A large percentage of the colonies are remote and difficult of access. The Ainsdale site on the S. Lancashire coast is on private ground on the boundary of the Ainsdale and Southport golf links and is a recent find of c. 1980. Its position partly within the golf course makes it locally safe. It is 1-5 km from high water on the coast and has a railway line and another golf links between it and the sea. Two former sites of Corynephorus canescens, at Formby, 6 km south of Ainsdale, have been lost by erosion, but at Ainsdale, accretion is now active and rabbits are contributing to the sand accretion requirements of Corynephorus canescens. Here the site isc. 100 X 4 m where our grass is dominant on two fixed dunes and has only Hypochaeris radicata and Ornithopus perpusillus as constant species. It is described by Savidge et al. (1963) as native. They record that there were formerly two sites between Formby and Freshfield in 1930, of which one became extinct in 1937 and the other was subsequently tidally eroded. The site at Hedderwick near Dunbar is a recent discovery of 1986 by Anna Younger and is considered an introduction. The area is a raised beach of blown sand with a fairly high proportion of shells which is reflected in the soil analysis. The site has been subject to much disturbance in the recent past from sand extraction to use as a tank training ground. The plants were found in two dune bays which were surveyed in 1995 and found to have at least 275 clumps in the main bay of c. 5 m diameter and 35 clumps in the other bay of 1 m diameter. The site is enclosed by Pteridium aquilinum with some Ammophila arenaria colonising the east end. If the advances of these taxa can be held in check, the site is probably safeguarded. The constant species at Hedderwick are different as here we find Arenaria serpyllifolia, Erodium cicutarium, Sedum acre and Thymus polytrichus. One area is being colonised with a low mossy turf of Brachythecium sp., Tortula spp. and including a Peltigera sp. _ At Lossiemouth, Morayshire is another introduced site, which is in a disused gravel pit, where 2000 plus plants occupy an area of c. 20 X 14 m. The plants are on a steep bank with Pilosella DISTRIBUTION AND STATUS OF CORYNEPHORUS CANESCENS 45 officinarum and Hypochaeris radicata. Another colony by a caravan site has a similar number of plants but spread over a wider area; those on sandy heath have Erica cinerea, Calluna vulgaris and Lotus corniculatus as companion species. The single plant of Corynephorus canescens found at Toigal by Morar, West Inverness-shire (Trist 1992) was on low dunes of highly mobile white sand on the shore of the tidal R. Morar. It was outnumbered by its constant species which included a single plant each of Aira praecox, Rumex acetosella subsp. acetosella and Carex arenaria. A few paces away was a small dune with a solitary plant of Calluna vulgaris. This was once a thriving colony which may have been introduced to this site. It was first found by F. Townsend in 1895 and now only a single plant bears witness. The Jersey, Channel Islands, sites are all found in the south-west of the island on coastal dunes and where some of the colonies are extensive and would experience salt spray from high tide. The dunes at St Ouen are relatively flat compared with those of Les Quennevais which attain a little height where the sands are deep and are the largest systems on the island. These latter calcareous dunes of broken down shells are remarkably rich in species which include Bupleurum fruticosum, Hornungia petraea, Rosa pimpinellifolia and Viola kitaibeliana with large colonies of Corynephorus canescens. On the dunes at Les Quennevais and L’Ouaisné, Lagurus ovatus and Poa bulbosa are frequent. THE INLAND SITES : One of the Corynephorus canescens sites in W. Suffolk is on Lakenheath Airfield where there are two stations. One is extensive and spread over an area of c. 0-4 ha which overlaps the parishes of Lakenheath and Wangford. The other, in the parish of Lakenheath, is c. 0-2 ha. The former has a population exceeding 5000 plants and the latter c. 500 plants. The site was formerly part of Lakenheath and Wangford Warrens which lie to the west of the A1065 road. It had been sheep grazed from early medieval times to 1942 when it was requisitioned by the Ministry of Defence. It was also an extensive rabbit warren for 700 years and there is some evidence of former cultivation (Crompton 1975). This western part of Lakenheath Warren was occupied by a series of dunes up to 1 km in width in the 1930s, with some dunes up to 6 m in height (Watt 1936). Following requisition, the dunes were destroyed and the entire area bulldozed flat and prepared for an airfield. The C. canescens recorded here is a relict of the find by Druce (1883, OXF) and the rediscovery by Crompton (1980). These airfield sites are mown twice monthly in the season and, in spite of this record of disturbance, a number of species survive. The large area contains Agrostis vinealis, Crepis capillaris, Campanula rotundifolia, Galium verum and Silene otites, while the smaller areas contribute Coronopus squamatus, Festuca ovina subsp. hirtula and a large colony of Thymus serpyllum. The other site in W. Suffolk is also in the adjoining parish of Wangford. It lies 2 km north-east of the Airfield sites and 60 years ago would have been part of the huge dune system on the warren. This site is fully described under conservation at Wangford Glebe. At Kinver Edge, which lies south-west of Kinver, Staffordshire, there are two sites of C. canescens which are c. 170 m apart and separated by an area of gorse, scrub oak and birch. The soil is a pebbly red sandstone with pH 6-7 which lies over Triassic Bunter Pebble Beds. The plants were found on a gentle slope of open semi-stable sand at 130 m. The constant species include seven mosses and three lichens with 24 higher plants which include Calluna vulgaris, Deschampsia cespitosa, Erica cinerea, Luzula campestris and Teucrium scorodonia. The site has been the property of the National Trust since 1917 and is c. 80 ha in extent. Although it is near to dwellings there is no encroaching urbanisation. The conservation of these two sites is related to passing walkers who are probably mainly responsible for sand compaction and so preventing accretion about the C. canescens. The population of c. 500 plants has remained fairly constant since 1977. The site is wardened and protective action is taken. | The Worcester sites consist of five colonies separated along the Severn Valley Railway, south of Devil’s Spittleful where three of the sites are on the track side with two small colonies and a third site reported in 1991 to have an “abundant colony”. Devil’s Spittleful is under the care of the local District Council as a Nature Reserve. Two other sites are found on the embankment where, in 1989, a field meeting recorded “several plants’’, and a “large local area’”’ reported in 1995. Another is located on Burlish Top which is c. 500 m south of the railway, where it was reported “‘locally 46 P. J. O. TRIST common” in 1989. The sites are within an extensive area of sand covering an area 2 X 2 km and provide sandy tracks, heath and embankment habitats. The C. canescens plants spread along the rail track have originated from the adjacent sandy heath. It seems clear that these several small locations of C. canescens represent a relict of a one-time large open area of sand which over time has become fragmented and buried under the neighbouring towns. Silene conica, which is now largely confined to East Anglia, has survived on Hartlebury Common since its discovery in 1900 and this site is within a few kms of the Devil’s Spittleful C. canescens locations. Teesdalia nudicaulis is still frequent, but Rosa pimpinellifolia and Erodium maritimum found on these sites in recent years are now extinct. Our plant is carefully monitored by the vice-county recorder. To give some interpretation of the present day inland and coastal habitats of C. canescens, we may consider the maritime sand habitats of the grass Elytrigia repens subsp. arenosa, which has a defined north-western European distribution which includes the east and south coasts of Britain (Trist 1995); it also occurs on heathland and sandy arable land in the Mainz area of Germany. On this extensive inland sand area many of the local taxa are also common to the sands of the W. Suffolk Breckland; both contain C. canescens and more than 20 other grasses (Hecker 1987). With the exception of those sites which are known or believed to be introductions, we may regard the W. Suffolk, Staffordshire and Worcestershire sites as relict areas which have resisted erosion. THE DISTRIBUTION OF C. CANESCENS IN BRITAIN AND THE CHANNEL ISLANDS E. Suffolk, v.c. 25. Dune at rear of beach, Minsmere Haven. TM/479.674, 18 July 1992, P. J. O. Trist; beach near Benacre Broad, TM/534.841, July 1995, P. Lawson and Y. Leonard; the shore, Kessingland Beach, TM/535.857, 7 June 1991, P. J. O. Trist. W. Suffolk, v.c. 26. Fixed dunes, Wangford Glebe, Brandon, TL/757.843, 15 July 1994, P. J. O. Trist; heathland west of main entrance to Lakenheath Airfield, TL/745.3809 and adjacent to runway, south of Wangford Farm buildings, TL/748.827—749.828, 18 July 1995, Y. Leonard and P. J. O. Trist. E. Norfolk, v.c. 27. Coastal dunes, North Denes, Gt Yarmouth, TG/533.090 and uninterrupted on dunes north to the Lifeboat station, Caister-on-Sea, TG/528.118 and continuing north and through the Caister golf course to TG/527.122, 5 September 1995, A. Bull; mobile dunes, Winterton-on- Sea, TG/502.190, 10 September 1992, P. J. O. Trist; north to Warren Farm, north-east of Horsey Corner, TG/460.247, 5 September 1995, A. Bull; grey dunes, Blakeney Point, TG/000.465—008.465, 4 September 1994, A. Bull. W. Norfolk, v.c. 28. Coastal dunes, Holme next the Sea, TF/705.445, 13 June 1995, G. Beckett and R.. j..O. Amst. Worc., v.c. 37. Severn Valley Railway track adjacent to Devil’s Spittleful, south-east of Wyre Forest, near Kidderminster: a. railway track side, Wribbenhall Junction, SO/798.749, 6 August 1991, R. Maskew and W. A. Thompson; b. railway trackside, SO/809.745—810.745, 6 August 1991, W. A. Thompson; c. railway embankment, SO/805.746, 30 July 1989, Worcs. Flora meeting; d. railway track side, SO/826.750, 29 June 1995, R. Maskew and W. A. Thompson; e. railway embankment, SO/832.752, 29 June 1995, R. Maskew and W. A. Thompson; sandy heath on Burlish Top, south-east of Droppingwells Farm, south-west of Kidderminster, SO/810.739-811.741, 30 July 1989, Worcs. Flora meeting. Staffs., v.c. 39. Sandy track side, Kinver Edge, south-west of Stourbridge, SO/834.827—834.825, October 1995, B. R. Fowler. S. Lancs., v.c. 59. On sand boundary with Ainsdale and Southport golf links, SD/318.127, 28 June 1994, V. Gordon. E. Lothian, v.c. 82. Sandy turf by the estuary at Hedderwick Hill Plantation, John Muir Country Park, west of Dunbar, NT/642.789, 3 September 1995, H. Jackson and A. Silverside. Moray, v.c. 95. Disused gravel pit, south of Lossiemouth, NJ/231.694, 24 July 1992, J. Edelsten and D. Law; caravan site on coast east of Lossiemouth, NJ/239.699, 24 July 1992, J. Edlesten and D. Law. Westerness, v.c. 97. Mobile dunes, Toigal by the estuary of the R. Morar, near Arisaig, NM/ 674.922, 24 July 1991, A. Slack, E. Norman and P. J. O. Trist. DISTRIBUTION AND STATUS OF CORYNEPHORUS CANESCENS 47 Channel Isles, v.c. 113. Jersey, coast dunes, Ouaisné, St Brelades Bay; coast dunes, La Pulente, St Ouen’s Bay; coast dunes, La Carriére, St Ouen’s Bay; higher interior dunes, Les Quennevais. All sites reported June 1995, J. Banks. GENERAL DISTRIBUTION Britain to S. Baltic and to S. Portugal, N. Italy and across to Central Ukraine, local in the eastern part of its range. ACKNOWLEDGMENTS I thank Gillian Beckett, Nick Gibbons and Yvonne Leonard for assistance in the field. I also thank Joan Banks, Alec Bull, John Edelsten, Bryan Fowler, Vera Gordon, Helen Jackson and Roger Maskew for site information. I am grateful to Gigi Crompton for information on the history of Lakenheath Warren, to Brian Davies for soil analysis, to Dr Max Walters and Derek Wells for assistance with the manuscript. I also acknowledge the interest of Derek Moore, the Director of the Suffolk Wildlife Trust and the Breckland Project officer. REFERENCES Crompton, G. (1975). The historical ecology of Lakenheath Warren in Suffolk, England: A case study. Biological conservation 8: 299-313. Crompton, G. (1980). Corynephorus canescens (L.) P. Beauv., in Plant Records. Watsonia 14: 200. Hecker, U. (1987). Corynephorus canescens (L.) P. Beauv., in Die Farn-und Blitenpflanzen des Mainzer Sandes. Mainzer Naturwissenschaftliche Archiv 25: 88. Hinp, W. M. (1889). The flora of Suffolk. Gurney & Jackson, London. MarsHALL, J. K. (1967). Corynephorus canescens (L.) P. Beauv, in Biological Flora of the British Isles. Journal of ecology 55: 207-220. SavipDGE, J. P., HEywoop, V. H. & Gorpon, V. eds. (1963). Travis’s Flora of South Lancashire, p. 137. Liverpool Botanical Society, Liverpool. SmitH, A. J. E. (1980). The moss Flora of Britain and Ireland. Cambridge University Press, Cambridge. Stace, C. A. (1997). New Flora of the British Isles, 2nd ed. Cambridge University Press, Cambridge. Trist, P. J. O. (1971). Corynephorus canescens (L.) P. Beauv. Watsonia 8: 402. Trist, P. J. O. (1980). Corynephorus canescens (L.) P. Beauv. in W. Suffolk, v.c. 26. Watsonia 13: 61-62. Trist, P. J. O. (1992). Corynephorus canescens (L.) P. Beauv. in Scotland. B.S.B.I. Scottish newsletter 14: 13— 15: Trist, P. J. O. (1994). Corynephorus canescens (L.) P. Beauv., in STEwarT, A., PEARMAN, D. A. & PRESTON, C. D. , eds. Scarce plants in Britain, p. 122. J.N.C.C., Peterborough. ‘ GRisr, P. J. O. (1995). Elytrigia repens (L.) Desv. ex Nevski subsp. arenosa 1 (Spenner) A. Love (Poaceae) in north- western Europe. Watsonia 20: 385-390. Warr, A. S. (1936). Studies in the ecology of Breckland, 1. Journal of sbes 24: 126-131. WRIGHT, T. (1668). Sandstorm. Philosophical transactions of the Royal Society 37: 722-725. (Accepted July 1997) (EDITOR’S NOTE: This paper was submitted by Derek Wells to Watsonia following the death of John Trist in June 1996. Derek took responsibility for ensuring that the paper was complete and for attending to editorial matters through the publication process.) ARAMA AORN aan t2,.oseelasl sdhwomrab roepsaetelintor ‘ rely on <') (iat Eheao, EVOL RTH | (ocateenn, ee sAn erthttiy ciation aor ae o) any: snet oath ye aciincete i whieh ever toe: Lee Coane yes Sp CD i THT ‘chababhaneahanset 4: Ehate ET * AAS Cl EMI Pa pa EE CF NRW Fea ore (Ute rae nad ite jh +m ate HO) TOR? MT Ee ; ? . yeeres m4 Na irhabed sete nokia, bse faa os “0 rie er via Otte whi i eae: and ni Fig, oe Ua i i aa fai peti Hitamest J TbaE, ish net. I rmbt: | A | enebetip meqedgnk alekt yields min gable ints aan oppriaind Vol shh mui reeaiting rekon genee gO) in te aihene Seek cobvst Hetsieitios ‘oie tasiord Sncblood ol ina eh Bae eek 5 be | ee ee rT Bay are 7 ve AS e Pom rf aie erie nk “fateh ne lvh eC ashi aro dogth tn marae, iaioghatvonitor col edheqiany gt a aint aaryoGhedy ; pa OAS Heid et -, et Apt) y bat bree’ pet, aint oksiainiaad Hin tReeoos naive » 4 oy be Dae Sas biciatl ‘cost it er ok PEARSE i: . ber: ant BY 4 fat ene 1...) Severe rane ay t suis) ae ae yl me aint Shay: ‘ 1 im vay - eo Wey aN ‘ats Jara tang (eu, daa. ar a AN Joctner) bates - a re Lp, WES B53 My ee : a tare be) ad i Liber ' Deh) By eh ae ae ad r*, vaio ie My eu'4 ie “ithe Bt waeatl 44 4 ee ae eset «aad igi ¥ ghee. ‘Gates Davis ane eb Ves nee ey bad .yradyipinga he pein F (1) vee UR cs et SOL Yd 2619 ‘baidlveae rae WT CNTR ein ot Wy io eed Paar} any ahr Lo PU ee, A EE WievouNeOA sh) Ga ye a i ng silica aii ial Degen ESS, Juh 41% be Winey, ove ie: A ct WAY CT Viodory) ane Nis i, * ks ey geet Lt} 27 GanOne A, Ae 5 ¥, AR fa! % A ‘uy LEE: Pot ee ‘ Ky ie EY tio i bare aby fi PEE tho) ey, Re GRa ty (cenit WORST AneEe ie nite wh, Saat.) Lie ws cue, WRG Ae BR nnchehelt ew te sf . ' — Va Voth A) S ES sone Awe: +7 at a i ‘we rer < a: Mis mal Vit hg m a \va LORS WEES WA iy pety dindbbins +: fy Ati: " re 4 ie 34 « ; %, vee 7s ‘ Sas hs 4 Gi Aik tina f Pie on “tity sick ot Soar hie ear ATS ie are ‘ eu) ‘tke 3 Pe ay eit | hoa ae ars Ve Lace, Rees ye SS , ean x 4, f et EP : ee) te fi i’ 4 itt hal . i js } / in { i oorhinned ad yaa ; Fe 1y SV Th H at ally Aerred iy Diy ieg tot Bes, asin cee Tate als vat aie Ve Poise May Saree (.exsocney 0 lo (igeb sii soawatlo . 4 id it id “oe nod Le via) 4 as © ‘4 - ” > >. Watsonia 22: 49-60 (1998) 49 Hybrids of Epilobium brunnescens (Cockayne) Raven & Engelhorn (Onagraceae) and their occurrence in the British Isles G. D. KITCHENER Crown Villa, Otford Lane, Halstead, Sevenoaks, Kent, TN14 7EA and D. R. MCKEAN Royal Botanic Garden, 20a Inverleith Row, Edinburgh, EH3 5LR ABSTRACT Data are presented on the occurrence of five new hybrids of Epilobium brunnescens (Cockayne) Raven & Engelhorn (Onagraceae) growing wild in the British Isles. Descriptions are given of E. brunnescens X ciliatum, E. brunnescens X lanceolatum, E. brunnescens X montanum, E. brunnescens X obscurum and E. brunnescens X palustre. These are named respectively as Epilobium X< brunnatum Kitchener & McKean hybr. nov., E. X cornubiense Kitchener & McKean hybr. nov., E. X confusilobum Kitchener & McKean hybr. nov., E. X obscurescens Kitchener & McKean hybr. nov. and E. x chateri Kitchener & McKean hybr. nov. Records of E. brunnescens X ciliatum and E. brunnescens X montanum in the British Isles are re-evaluated. KeEyworps: hybridization, willowherbs, New Zealand. INTRODUCTION Epilobium brunnescens (Cockayne) Raven & Engelhorn (New Zealand Willowherb) (Onagraceae) has long been naturalized in the British Isles. It was first recorded in 1904 at Craigmillar, Edinburgh, as a garden weed (Fraser 1905). It has been suggested that E. brunnescens was probably introduced into Britain for planting in rock gardens where it may have become a rampant weed. It was also brought in with wrappings and soil attached to other New Zealand plants. Its gradual spread accelerated from the 1930s (Davey 1953, 1961) and it now has a substantial presence in those parts of the British Isles whose high rainfall and humidity approximate best to its New Zealand habitats, where it is “especially common on shingly riverbeds in areas of high rainfall’”’ (Webb et al. 1988). Its corresponding habitats in the British Isles are in the west and north or in upland areas: moist open areas, on gravel, gritty or stony soils, streamsides, ditches, paths, screes, damp stone walls and banks. Outside Britain, Ireland and its native home, E. brunnescens is also reported from Norway as a rarity and mainly as a garden weed. Although originally reported there in 1931 (Davy 1961) it has not become invasive as in Britain and Ireland. Although Epilobium ciliatum, E. montanum, E. obscurum and E. parviflorum (but not E. lanceolatum or E. palustre) have been, or are, found in New Zealand, only E. ciliatum has, as yet, been reported as having hybridized in the wild there with E. brunnescens (Raven & Raven 1976). The nomenclature used here follows that of Stace (1991) and Kent (1992). ARTIFICIAL HYBRIDIZATION Some data are available as to the synthesizing of hybrids between E. brunnescens and Epilobium species present in Britain. Brockie (1966) succeeded in crossing E. brunnescens (sub nomine E. pedunculare) with E. ciliatum (sub nomine E. erectum). From the progeny, an F, generation was raised. Brockie (1970) also reports an attempt by V. Thakur in the 1960s to cross E. brunnescens 50 G. D. KITCHENER AND D. R. MCKEAN (sub nomine E. nerterioides) with E. montanum in both directions, but with failure to set seed. It should not be assumed that this is evidence of a significant barrier against cross-fertilization, for it is not clear that enough attempts were made to effect good fertilization. Raven (1972) considered the failure of good seed set reported by Thakur and Brockie (1970) and concluded that there was no evidence for failure of good set as a barrier to hybridization between Australasian species of Epilobium. He also considered that there was no definite evidence for the operation of such a barrier between any two species of sect. Epilobium from anywhere in the world, with the possible exception of short styled small flowered species being incompatible with long styled large flowered species. THE IRISH RECORDS Northern Ireland accounts for the first records of E. brunnescens hybrids in the British Isles, but these have been the source of considerable confusion. Scannell & Synnott (1987) mention E. brunnescens X ciliatum and E. brunnescens X montanum as being present in v.c. H39. The former record is elaborated by Hackney (1992), who refers to E. brunnescens X ciliatum as being “abundant with both parents’. The latter record relates to the specimen which resulted in “E. brunnescens X ?E. montanum’ (sic) being credited to the flora of the British Isles by Stace (1991). We have examined the relevant specimens (BEL) and it appears to us that the specimen determined as E. brunnescens X ciliatum is E. brunnescens X obscurum, together with a number of other plants (E) from the same location; the specimen determined as E. brunnescens X montanum is probably E. brunnescens X ciliatum. All these records related to finds at the spoil tips of a quarry at Magheramorne, near the shore of Larne Lough in Co. Antrim (v.c. H39). A series of hybrid specimens were collected by D. Ledsham in 1980, and these now form three sheets held at E. Although sent for identification in 1981, they were unfortunately mislaid until 1994, when they were identified by one of us (D.R.M.) as E. brunnescens X obscurum, this being confirmed by T. D. Pennington. Two subsequent collections were made from the same locality. The first, in 1981, was made by D. Ledsham and D. Getty, tentatively identified by C. A. Stace as E. brunnescens X montanum and so confirmed by P. H. Raven in 1982 (specimen held at BEL). An unqualified determination by Raven must carry considerable weight of authority, but this is difficult material. The flowers of such a cross may be expected to bear ‘“‘confused”’ stigmas, i.e. almost clavate, but with traces of lobing (resulting from crosses between parents with four-lobed and clavate stigmas). The absence of these is not conclusive, because stigma-shape may vary on the same plant, but examination in 1996 of the most developed flower on this material (H8544) showed a clavate stigma. This suggests either E. obscurum or E. ciliatum as the other parent; the latter is more likely on the basis of stem and capsule hair characters. | Further evidence lies in the fact that E. montanum was altogether absent from the site, whereas E. ciliatum was frequent (fide D. Getty; P. Hackney, pers. comm., 1993). On the strength of this, E. brunnescens X ciliatum is given in the current local Flora (Hackney 1992), instead of E. brunnescens x montanum. The two different determinations of the same single specimen, however, have resulted in both hybrid taxa being recorded in Scannell & Synnott (1987). The second collection (BEL) was made in 1983 by R. Piper and was, following the redetermi- nation of the 1981 collection, tentatively determined by Getty and Hackney as E. brunnescens X ciliatum. The very limited presence of glandular hairs in the upper parts, however, points to this material being E. brunnescens X obscurum, as with the bulk of the specimens from this site. A further Irish discovery of E. brunnescens X obscurum was made by J. McNeill (det. D.R.M.) in 1992. This is represented by a fragmentary specimen at E, part of a stem curving from a prostrate to semi-erect position. It was found in the Glenelly valley in the Sperrin mountains in Co. Tyrone (v.c. H36) in the company of E. brunnescens, E. obscurum, E. montanum and E. palustre. THE BRITISH RECORDS E. brunnescens X ciliatum was first recorded in Britain during 1995, in v.cc. 1, 2 and 44. In June - 1995, A. O. Chater (pers. comm., 1995) noted about 100 plants on a gravelly and stony quarry slope EPILOBIUM BRUNNESCENS HYBRIDS IN THE BRITISH ISLES 51 near Llansawel, Carms. (v.c. 44), three specimens from this population being confirmed as such by T. D. Pennington. In July 1995, about 20 plants were located by D. T. Holyoak (pers. comm., 1995) on mine waste at Wheal Maid, near United Downs, West Cornwall (v.c. 1). A further plant was found by R. J. Murphy (pers. comm., 1995) and H. Meredith at the end of that month on china clay gravels at a disused tip at Goonamarris near Nanpean, East Cornwall (v.c. 2). The United Downs locality was re-investigated by D. T. Holyoak (pers. comm., 1996) in July 1996, and while E. brunnescens X ciliatum was present in its usual small, red, semi-prostrate form, there were two near-erect plants, much larger and well branched. These are discussed in the comparison below. In the same month, he discovered six more specimens of this hybrid on mine spoil at Wheal Busy, West Cornwall (v.c. 1) — these have not been examined by us. Also in July 1996 one of us (G.D.K.) found six further plants, again on china clay gravels: three south of Burngullow Common, and the other three at Carclaze Downs pit, both locations being near St Austell, East Cornwall (v.c. 2). E. brunnescens X lanceolatum was first discovered by D. T. Holyoak in July 1995 as a single small plant in the same United Downs location, being also in the company of E. brunnescens, E. ciliatum, E. hirsutum, E. lanceolatum, E. montanum, E. obscurum, E. parviflorum, E. ciliatum X montanum and E. ciliatum X parviflorum. The presence of E. lanceolatum, and hence its potential as a parent, was not recognised until the next year and so, unfortunately, this record was mistakenly published by one of us (Kitchener 1996) as E. brunnescens X montanum. The southern distribution of E. lanceolatum in Britain renders it unlikely that E. brunnescens X lanceolatum would be found other than in Cornwall, Devon or South Wales. A further specimen was located by D. T. Holyoak in July 1996, again at United Downs, West Cornwall (v.c. 1). E. brunnescens X montanum, discounting the supposed Irish records mentioned above, was first found in the British Isles in July 1996 by one of us (G.D.K.) with R. J. Murphy and M. & A. Atkinson, growing on imported granite aggregate and rubble at Clicker quarry near Liskeard, East Cornwall (v.c. 2). E. brunnescens X obscurum was first recorded in Britain in East Cornwall (v.c. 2), again from Clicker quarry, where R. J. Murphy discovered it on the occasion, and with the people, mentioned above (M. Atkinson having observed potential E. brunnescens hybrids there in the previous year). Also in July 1996 in v.c. 2, one of us (G.D.K.) found three plants on china clay gravels: two at Carclaze Downs pit, and a further one in a china clay pit north east of Stenalees. E. brunnescens X palustre was first recorded in the British Isles by A. O. Chater in August 1995, growing on damp acidic shaley soil by a forestry road in a conifer forest near Hardro, Llyn Brianne, Cards. (v.c. 46). E. palustre was growing nearby, and the specimen has been confirmed by T. D. Pennington. DESCRIPTIONS OF TAXA Material from all the collections listed above has been examined by us in the compilation of the following descriptions unless otherwise stated. Epilobium < brunnatum Kitchener & McKean, hybr. nov. (Epilobium brunnescens (Cockayne) Raven & Engelhorn X E. ciliatum Raf.) (Fig. 1) Hybrida inter Epilobium brunnescens (Cockayne) Raven & Engelhorn et E. ciliatum Raf., characteribus inter parentes variantibus; planta semi-prostrata, stigmatibus clavatis, seminibus praecipue sterilibus, eorum pagina plerumque tuberculato-costata, et in ovario pilis numerosis glandulis praeditis, aliquot pilorum ad pedicellum subtentum extensorum. - Herb with prostrate leafy runners, with stems much branched, curving up to an erect position, 8— 20(—30) cm high and little branched above, except in cultivation, when growth habit is more erect and branched. Stems with two lines of short crisped hairs, descending from the nodes, and beginning © between a pair of petioles and descending to the next node; the hairs become generally scattered on the stem upper parts and may be absent near the base. Leaves mainly opposite but largely alternate in the upper parts, green on younger growth, especially the upper surface; sessile or with petioles not exceeding 1-5 mm; cauline leaves (0-5—)0-8-1-2(—2:3) x (0-3—)0-40-6(—1-1) cm, average ratio of 52 G. D. KITCHENER AND D. R. McKEAN length to breadth 1:0-4-0-5; ovate-lanceolate to elliptic with a few obscure teeth, glabrous except for marginal pubescence; leaves on runners similar, but tending to more elliptic shape. Sepals lanceolate, 2:-5—3-0 x 0-7-1-2 mm, bearing crisped and glandular hairs. Flowers very pale pink, rarely darker c. 9 mm diameter, with petals 5-6 mm long. Four long stamens projecting to or just below top of stigma, their anthers being 0-55—0-85 mm long; four short stamens extending as far as stigma base. Ovary 0-7—1-8 cm, covered with dense patent glandular and crisped eglandular hairs, the glandular hairs extending to, but comparatively infrequent on, the pedicels. Stigma entire, clavate, 1-2-1-9 mm; style longer than stigma, erect, white, 1-3—2-8 mm. Capsules 1-2—3-0 cm long, with hairs as for ovary, sometimes twisted and shrivelled, mostly sterile, containing shrunken abortive seeds 0-34—0-47 mm long; also contained are a few larger seeds, either malformed (0-6-0-7 mm) or fully formed and fertile (0-8—0-9 mm) with rows of tubercules, often with the surface texture tending towards the ridged rows of E. ciliatum. The larger seeds may bear a short neck or appendage at the point of attachment of the coma. Hotortypus: West Cornwall, v.c. 1, Wheal Maid, near United Downs, grid reference SW/745.420, on mine waste, alt. 100 m, 5 July 1995, D. T. Holyoak (E). A wild population of this hybrid was seen by P. H. and T. E. Raven in 1969 or 1970 in New Zealand at the Whangaehu River, and they also note a collection from Canterbury, New Zealand in 1919 which is probably the same hybrid (Raven & Raven 1976). Epilobium X< cornubiense Kitchener & McKean, hybr. nov. (Epilobium brunnescens (Cockayne) Raven & Engelhorn x E. lanceolatum Sebast. & Mauri) Hybrida inter Epilobium brunnescens (Cockayne) Raven & Engelhorn et E. lanceolatum Sebast. & Mauri, characteribus inter parentes variantibus; planta semi-prostrata stigmata lobis confusis ferens, caulibus pilis crispatis brevibus uniformiter obtectis, et floribus in colore ut minimum aliquot spectri colorum E. lanceolati exhibentibus. Herb with stems curving up to erect position, 6-5—25 cm high, little branched above except for larger stems, and with a fairly uniform covering of short crisped hairs. Leaves opposite, largely alternate on infloresence, reddish, sessile or with petioles not exceeding 1 mm; cauline leaves 0-5— 1-1(-1-5) x 0-7(-0-9) cm, average ratio of length to breadth 1:0-57—0-63; basal leaves crowded; leaves ovate to broadly elliptic with a few obscure teeth, glabrous except for marginal pubescence and a few hairs on the underside midrib. Sepals lanceolate, 2-8-3-5 x 1-9-1-4 mm, bearing short crisped and patent glandular hairs. Flowers white, pale pink or very pale pink, and the colour may vary on the same plant, but not on the same flower; with petals c. 5 mm. Four long stamens extending to, or just below stigma, their anthers being c. 0-6 X 0-35 mm; four short stamens well below stigma. Ovary 1-2—1-9 cm, covered with many short crisped hairs and some patent glandular ones, a few of the latter descending to the pedicels; hair coverage of pedicels less dense than that of upper stem (in contrast to E. X brunnatum). Stigmas variable, with confused partial lobing 0-8—1-0 mm; style 3-5—4 mm. Capsule 2-2-3-1 cm, mostly sterile, containing shrunken abortive seeds 0-35— 0-5 mm long, their surface with rows of platelets, sometimes reticulate; seeds occasionally fully formed, cylindrical, 0-65—0-95 mm with rows of tubercles. Ho .orypus: West Cornwall, v.c. 1, Wheal Maid, near United Downs, grid reference SW/745.420, on mine waste, alt. 100 m, 12 July 1995, D. T. Holyoak (E). Epilobium xX confusilobum Kitchener & McKean, hybr. nov. (Epilobium brunnescens (Cockayne) Raven & Engelhorn x E. montanum L.) Hybrida inter Epilobium brunnescens (Cockayne) Raven & Engelhorn et E. montanum L., characteribus inter parentes variantibus; planta semi-prostrata, stigmata lobis confusis ferens, caulibus pilis crispatis brevibus uniformiter obtectis, et floribus coloris pallide purpureorosei. Herb with stems branched at ground level, creeping at first, and then curving up to erect position, 5—10 cm high, scarcely branched above, and with a fairly uniform covering of short crisped hairs, diminishing in upper parts, and with occasional glandular hairs. Leaves opposite, largely alternate on inflorescence, pubescent petioles not exceeding 1(—1-5) mm; cauline leaves 0-5—1-0 x 0-25-0-5 cm, average ratio of length to breadth 1:0-45; basal leaves crowded; leaves ovate to broadly elliptic with a few obscure teeth; short crisped hairs on underside midrib and veins, with a light scattering elsewhere beneath, pubescent on margins; upper leaf surface varying from pubescent (especially on - young growth) to glabrescent. Sepals lanceolate, 1-45—2-90 x 0-73-1-02 mm, with a scattering of EPILOBIUM BRUNNESCENS HYBRIDS IN THE BRITISH ISLES 53 CZ aN: . PEs op > ie a }—+4 NS Ossie lem pore / Figure 1. Epilobium xX brunnatum. A. plant; B. mid-stem; C. flower (half cut away); D. sterile seed; E. fertile seed; F. transverse section of ovary. A glandular hairs. Flowers pale purplish pink, with petals c. 3-5 mm. Four long stamens projecting to top of stigma, their anthers being 0-4—0-5 x 0-3-0-35 mm; four short stamens extending to the mid- level or base of stigma. Ovary 0-7—1-25 cm, with lines of dense short glandular hairs, these hairs being scattered down the pedicel also. Stigma with confused partial lobing; style longer than the stigma. Capsule 1-6-2-6 cm, containing shrunken abortive seeds 0-35—0-45 mm long, surface flattish, marked in rows and sometimes bearing low tubercles; occasional larger seeds more fully formed, 0-65—0-95 mm, cylindrical with tubercular surface. 54 G. D. KITCHENER AND D. R. McKEAN Ficure 2. Epilobium X obscurescens. The stem on the right hand side is 23 cm long. Ho otypus: East Cornwall, v.c. 2, Clicker quarry, near Liskeard, grid reference SX/288.614, on rubble and imported granite gravel, alt. 100 m, 27 July 1995, coll. M. Atkinson (E). Epilobium xX obscurescens Kitchener & McKean, hybr. nov. (Epilobium brunnescens (Cockayne) Raven & Engelhorn xX E. obscurum Schreber) (Fig. 2) Hybrida inter Epilobium brunnescens (Cockayne) Raven & Engelhorn et E. obscurum Schreber, characteribus inter parentes variantibus; planta semi-prostrata, stigmatibus clavatis, et seminibus praecipue sterilibus, eorum pagina laevi et reticulatovenosa, et in ovario pilis nonnullis glandulosis praedito, eis ad pedicellos haud extensis. Herb, conspicuously reddish, with stems well branched mainly at ground level, creeping at first and then curving up to erect position, 8-23 cm high, sharply quadrangular and with raised lines running from one node to the next, with fairly sparse short crisped hairs, more numerous in upper parts. Leaves mainly opposite, but largely alternate on upper part of stems, ovate-lanceolate to elliptic with a few, 2—3(—5) obscure teeth, glabrous except for short marginal pubescence, sessile or with petioles not exceeding 1(—2) mm; cauline leaves 0-6—1-5 x 0-3-0-65 cm, average ratio of length to width 1:0-42—0-44. Sepals 3-3-5 x 1 mm, bearing crisped hairs. Flowers very pale pink, petals 5-6 EPILOBIUM BRUNNESCENS HYBRIDS IN THE BRITISH ISLES 55 mm. Ovary 1-3—2-2 cm, with frequent crisped hairs and occasional patent glandular hairs, the latter not extending down to pedicel. Stigma entire, clavate. Capsules 1-5—3-1 cm, densely curly hairy; seeds mostly sterile, 0-6—0-8 mm long when fully formed but the sterile ones only averaging 0-32 mm, surface smooth but with reticulate veining. Ho toryrus: Co. Antrim, v.c. H39, Magheramorne, grid reference J/43.98—44.98, on spoil tips of quarry, 1 July 1980, D. Ledsham, s.n., bar-code no. 33089 (E); isotypus (BEL). The specific name is spelt obscurescens rather than obscurascens because it is an epithet derived from part of each of the parental epithets. Epilobium X chateri Kitchener & McKean, hybr. nov. (Epilobium brunnescens (Cockayne) Raven & Engelhorn x E. palustre L.) (Fig. 3) Hybrida inter Epilobium brunnescens (Cockayne) Raven & Engelhorn et E. palustre L.., characteribus inter parentes variantibus; planta semi-prostrata stigmatibus clavatis, et seminibus praecipue sterilibus sed interdum fertilibus tum plus quam 1 mm longis. Herb with prostrate leafy runners, from which arise erect stems, 20-25 cm high and scarcely branched above, with 2 broad lines of short crisped hairs, descending from each node; the density of the hairs increasing in the upper parts of the stem. Leaves opposite, largely alternate on the flowering part of the stem, reddish, sessile or with petioles to 1 mm; cauline leaves 0-6-1-2 x 0-2-0-4 cm, average ratio of length to breadth 1:0-29; narrow, lanceolate with a few obscure teeth, glabrous except sometimes pubescent on margins or upper midrib. Leaves on runners similar, but smaller, up to 0:6 cm long, 0-2 cm wide and with more distinct petioles. Sepals c. 2-2 mm long, lanceolate and strigillose. Flowers very pale pink. Four long stamens projecting to lower part of stigma, their anthers being 0-43—0-50 mm long; four short stamens extending to style below. Ovary c. 1.3 cm, covered with short crisped and patent hairs, some glandular. Stigma entire, clavate, 0-95—1-76 mm long; style longer, erect, 1-42—2-37 mm. Capsules 1-5—3-3 cm long, with numerous crisped hairs and some patent glandular hairs; mostly sterile, containing shrivelled, abortive seeds 0-4 mm long; also contained are a few larger seeds 1-0—-1-3 mm long with tubercled surface, generally part collapsed longitudinally, but occasionally fully formed and fertile. The larger seeds bear a neck or appendage at the point of attachment of the coma. Ho ortypus: Cardiganshire, v.c. 46, south of Hadre, Llyn Brianne, grid reference SN/804.512, on damp, acidic shaley soil by Forestry Commission road, alt. 370m, 7 August 1995, A. O. Chater, s.n. (E); isotypus (NMW). COMPARISON OF TAXA Virtually all wild specimens were noticeable in the field as having red stems and leaves. Growth began as prostrate, often spreading radially, with flowering stems curving up to an erect or semi- erect position. That growth habit is less conspicuous in cultivation. A specimen of E. X brunnatum was grown on by A. O. Chater and assumed a sprawling, multi-branched erect habit. This was replicated by one of us (G.D.K.) in growing on both E. X brunnatum and E. X obscurescens. A similar result arose from cultivation of experimental hybrids made by Brockie between E. brunnescens and E. ciliatum, it being reported by Raven & Raven (1976) that plants grew into large mounds about 1 m across, with very little about them to remind one of E. brunnescens. The cultivation by one of us (G.D.K.) of F) specimens of E. X brunnatum and E. X chateri also gave analogous results, with plants demonstrating extensive scrambling growth. Reddening generally only occurred at a late stage, when plants were transferred outdoors: it may derive from stress or exposure. There have been a few instances where plants have been found in the wild with a more vigorous and nearly erect growth habit. One such was the 1981 Irish specimen mentioned above, to which the name of FE. brunnescens X montanum was initially applied, but which appears to be EF. xX brunnatum. This is a well-branched green specimen rising to 20 cm. Others are the two 1996 West Cornish E. X brunnatum plants discovered by D. T. Holyoak. These rise to 24 and 28 cm, bear leaves over twice as large as are normally seen in such hybrids, and in general seem much closer to the erect parent. More favourable growth conditions may be suspected, although these latter plants were apparently growing in stony mine waste. 56 G. D. KITCHENER AND D. R. MCKEAN st -— 7 Icom Fc > ta) A Y = Ie Ar 9 (= Y AP of 0 RS mal 9 a oN ¢ 5 wre’ * ie ka \ p AQRG ey i WW = =~ Ficure 3. Epilobium X chateri. A. plant; B. flower; C. fertile seed; D. sterile seed; E. transverse section of ovary; F. mid-stem just below node. The hybrids follow the erect species in bearing alternate leaves in the upper parts, instead of continuously opposite leaves with flowers borne individually in their axils, which is a characteristic peculiar to certain Australasian species of Epilobium, including E. brunnescens (Raven 1972). The wild specimens also show variable capsule length with very limited seed set, as is consistent with hybrid status. Their red colouring generally extends also to the sepals, which often bear an apical purple knob, as with E. brunnescens. The red sepals contrast with the greenish collar below. These sepal characteristics have not been repeated in each of the descriptions given above. EPILOBIUM BRUNNESCENS HYBRIDS IN THE BRITISH ISLES if] The morphological distinctness of prostrate EF. brunnescens is so much greater than the differences between the various erect British species, however, that FE. brunnescens is more readily identified as a parent than the other contributing species. In identifying the other parent, an analysis based on the species groupings put forward by Stace (1975) is of assistance. The initial grouping separates species with clavate stigmas and those with the stigmas four-lobed. As E. brunnescens has clavate stigmas, its hybrids with the former will carry similar stigmas, and its hybrids with the latter may be expected to bear variously intermediate stigmas (whose shapes are sometimes called “melted lobes” or “‘clenched fist’’). The holotype specimens of EF. x cornubiense and E. X confusilobum mentioned above indeed bear such intermediate stigmas. They are also distinct in carrying a fairly even spread of stem hairs which are short and crisped, representing the contribution of E. montanum or E. lanceolatum, as the case may be, and which are not conspicuously restricted to stem lines or bands. Some comment is perhaps due as regards the presence of glandular hairs on the upper parts of the specimens. Stace (1975) identifies E. montanum and E. lanceolatum as species which are not expected to contribute glandular hairs to a hybrid, and clearly they cannot be expected to do so to the same degree as, say, E. ciliatum. Many published descriptions do not mention glandular hairs in relation to these former species, although they do, nevertheless, occur. Exceptions are Stace (1991) and Haussknecht (1884): “Capsulis . . . junioribus tenuiter patentim glanduloso-pilosis” (E. montanum); and “‘Capsulis. . . pilis glandulosis brevissimis intermixtis obsitis; pedicellis glanduloso-puberulis” (E. lanceolatum). The most apparent distinction between E. X cornubiense and E. X confusilobum, where (as with the Cornish mining or quarry sites) one hybrid combination cannot be ruled out by the absence of one of the potential parents, lies in the flower colour. The corollas of E. lanceolatum vary in the course of growth from white through to light and deep shell pink, each flower on a particular plant being a uniform colour, but perhaps differing from others on the same plant at that time. It appears that these colour characteristics can be inherited by E. X cornubiense. The 1995 specimen carries two flowers which were noted in the field as being white, as distinct from the very pale pink that characterizes most E. brunnescens hybrids so far found; and also as distinct from EF. xX confusilobum, whose corollas, on the material so far seen, are purplish pink. The 1996 specimens of E. X cornubiense carried several flowers which were noted in the field as ranging from pale pink to very pale pink, and that record of variability on one plant (while each flower was of uniform colour) is a valuable observation. Hybrids with EF. palustre, E. ciliatum and E. obscurum all involve parents with clavate stigmas. Distinguishing features are as follows: E. X chateri is best recognised from leaf shape and (fertile) seed characteristics. E. palustre leaves are the narrowest of those of British species of Epilobium, and this character is not completely obscured by the broadly ovate to sub-orbicular leaf shape of E. brunnescens. The holotype hybrid specimen had a leaf length/breadth ratio of 1:0-29, significantly narrower than average measure- ments for material of E. x brunnatum (1:0-4-0-5), E. X cornubiense (1:0-57), E. X confusilobum (1:0-45) or E. X obscurescens (1:0-42-0-44). At the other end of the spectrum, the ratio for E. brunnescens is 1:0-7—0-85. E. palustre also has the largest seeds of British Epilobium species (1-6-1-8 mm), while E. brunnescens seeds, at 0-7 mm, are smaller than those of any of the native British species. The size of E. palustre seeds is reflected in the hybrid, whose fertile seeds (at 1-0-1-3 mm long) are longer than the fertile seeds of the other E. brunnescens crosses: E. X brunnatum (0-8-0-9 mm), E. X cornubiense (0:7-0-8 mm), E. X confusilobum (0-65—-0-8 mm) and E. x obscurescens (0-6-0-8 mm). The fertile E. < chateri seeds also inherit beak characteristics from E. palustre, and bear a neck or appendage at the comal end. E. X brunnatum and E. X obscurescens are very similar, but the former may be distinguished by the abundance of glandular hairs on the ovary deriving from E. ciliatum and descending at least down to the pedicels, such hairs being quite infrequent in the case of E. < obscurescens, and not present on the pedicels. The surface texture of the seeds also differs, with the sterile seeds of E. x obscurescens showing a pattern more reticulate than tuberculate (Fig. 4). It should be noted that the © surface texture of fertile seeds is not necessarily the same as that of the sterile ones. That difference may perhaps be interpreted as a failure of the sterile seeds to develop fully their surface features, so presenting a somewhat desiccated effect. So the low ridging of sterile E. x brunnatum seeds (reflecting the ridged patterning of E. ciliatum) may become more prominent and tubercled in the fertile seeds. The reticulation of sterile E. X obscurescens seeds does not represent a reticulate 58 G. D. KITCHENER AND D. R. McKEAN Ficure 4. Scanning electron micrographs showing surface texture of sterile seeds of E. X brunnatum (A,B) and E. X obscurescens (C,D). seeded parent. The surface of E. obscurum seeds is covered by narrow, wavy tubercules, and fertile E. X obscurescens seeds can repeat that appearance; so it seems likely that the reticulation of the sterile seeds also represents an undeveloped state of a tubercular surface. While it is often not easily observed, the presence of a short neck or appendage to the larger seeds, at the point of attachment of the coma, is an inheritance from E. ciliatum, and will clearly distinguish E. x brunnatum from E. xX obscurescens. STERILITY Numerous hybrids were present at the recorded locations in v.cc. 1, 44 and H39. This opens the possibility of second or subsequent generations of hybrids, whether through back-crossing or self fertilisation. The presence of occasional seeds, either fully formed or only in part malformed (generally through longitudinal collapse of one side) has been noted in the specimens examined. A sample capsule from the v.c. 1 material of E. x brunnatum yielded 85 small abortive seeds and six larger ones. If fertile, the latter such seeds might account for at least some plants where hybrid populations exist. Experimental sowing was undertaken in September 1995 using larger seeds taken from two of the v.c. 1 specimens of E. X brunnatum. Seeds were sown from each plant indoors, in small pots EPILOBIUM BRUNNESCENS HYBRIDS IN THE BRITISH ISLES 59 containing a mixture of John Innes No. 3 and ballast sand; moisture was conserved by enclosing each in transparent plastic bags. Seeds were not covered, since Brockie (1966), working with New Zealand Epilobium hybrids, considered that light appears necessary to ensure good germination. From 18 seeds taken from one plant, nine seedlings germinated after 8 to 14 days. From 14 seeds taken from the other, five seedlings germinated over 14 to 39 days. The F, seedlings were transplanted into individual pots and, with some losses in cultivation, six plants survived successfully into the next year. Germination of four F, seedlings was also achieved from material of E. x chateri. The F, plants themselves set occasional apparently fertile seeds, as with the F; generation; one plant of E. x brunnatum was, however, apparently totally sterile, with collapsed, malformed stigmas. Apart from the general differences in cultivated material mentioned above, the F, plants showed a degree of variation which, if replicated in the wild, would render it even more difficult in some cases to identify what taxa were involved. The progeny of EF. X chateri, for example, did not exhibit leaves quite as narrow as those from the original find (these wider leaved forms have been discounted in calculating the leaf length/breadth ratio given in the description of that hybrid). DISTRIBUTION AND HABITAT The occurrence of E. brunnescens hybrids in eleven localities in the British Isles up to 1996 (v.cc. 1, 2, 44, 46, H36, H39) provides limited scope for generalization. But data sufficient for mapping purposes have seldom been gathered in respect of any Epilobium hybrids — see Kitchener (1990), where it is concluded that “the real barrier to a fair assessment of the occurrence of willowherb hybrids has been lack of recognition, rather than scarcity’. Common to most of these localities, however, is an acid substrate, generally damp, open and of artificial origin; mining spoil appears particularly suitable. The importance of open, disturbed habitats (e.g. quarries and wasteland) has been emphasised by Stace (1975) as affording a range of suitable opportunities for the establishment of Epilobium hybrids. But, in order for these opportunities to be taken up, there are at least two other relevant factors: the relative genetic compatibility of potential parents, and the availability of those parents themselves. As regards the availability of the other parents, each of the species examined in this paper is readily to be found within the distributional range of E. brunnescens. As a limited measure of this — and taking into account the warning given that the data are not to be interpreted as showing comprehensive distribution — one may take the results of the B.S.B.I. sample survey under its Monitoring Scheme of 1987-8 (Palmer & Bratton 1995). E. brunnescens is recorded as present in 1987-8 in 125 of the sample squares, primarily in the damper western and northern parts and uplands of the British Isles. A high degree of coincidence is shown by the ubiquitous E. montanum (present in 90% of the same squares), by E. palustre (95%) and E. obscurum (82%). The coincidence level of E. ciliatum is less (46%), as this species began its spread from the south- east, and has consolidated in the drier parts of the British Isles, from which E. brunnescens is largely absent. It has, however, been appearing increasingly within the range of E. brunnescens, as demonstrated by Preston (1989), with substantial spread in Wales and Cornwall occurring in the period 1959-69, and in Scotland and Ireland through to 1986. There is no reason to suppose that the spread of E. ciliatum has yet ceased; it has since appeared in vice-counties additional to those stated by Preston for the period up to 1986 (personal observation, and Palmer & Bratton 1995). The increase of E. ciliatum may be relevant to the future occurrence of E. brunnescens hybrids if, as is possible, it is more ready to hybridize than other species, or such hybrids establish themselves with greater success. This is a matter for speculation, but some hybrid combinations of Epilobium are encountered more frequently in the field, and some species encountered more frequently as parents. Stace (1975) comments on this, and E. ciliatum, E. montanum and E. obscurum figure most prominently in the crosses which he regards as most frequently encountered. In our experience, within the range of E. ciliatum, it is the most frequently encountered parent in Epilobium hybrids. Further data are required before any assertion can be made about the relative frequency with which that species, in comparison with others, may hybridize with E. brunnescens. 60 G. D. KITCHENER AND D. R. MCKEAN CONCLUSION It is perhaps surprising that these taxa have not been recognized until relatively recently in the British Isles, despite their potential occurrence having been publicised by Stace (1976). The confused history of the Irish records and the temporary loss of several specimens has not assisted that situation. It seems at least possible that, with further fieldwork, E. brunnescens hybrids may be found not to be so rare as the hitherto limited number of records suggests. ACKNOWLEDGMENTS We would like to thank the following for provision of assistance, information or specimens: M. Atkinson, A. O. Chater, P. Hackney, D. T. Holyoak, D. Ledsham, J. McNeill, M. Mendum, R. J. Murphy, T. D. Pennington and Professor C. A. Stace. R. R. Mill kindly provided the Latin descriptions and advice. REFERENCES BrockIE, W. B. (1966). Artificial hybridisation of New Zealand species and varieties of Epilobium. New Zealand journal of botany 4: 366-391. BrockiE, W. B. (1970). Artificial hybridisation in Epilobium involving New Zealand, European and North American species. New Zealand journal of botany 8: 94-97. Davey, A. J. (1953). Epilobium pedunculare in Britain, in LousLey, J. E., ed. The changing flora of Britain, pp. 164-167. Botanical Society of the British Isles Conference Report, B.S.B.I., Arbroath. Davey, A. J. (1961). Biological Flora of the British Isles: Epilobium nerterioides A. Cunn. Journal of ecology 49: 753-759. Fraser, J. (1905). Alien plants near Edinburgh. Annals of Scottish natural history 53: 96-103. Hackney, P., ed. (1992). Stewart and Corry’s Flora of the North-east of Ireland, 3rd ed. Institute of Irish Studies, The Queen’s University of Belfast, Belfast. HAUSSKNECHT, C. (1884). Monographie der Gattung Epilobium. Gustav Fischer, Jena. JORGENSEN, P. M. (1992). New Zealand Willowherbs in Norway. Blyitia 50: 21-22. KITCHENER, G. D. (1990). Willow-herb hybrids and the great storm of 1987. Transactions of the Kent Field Club 11: 69-73. KITCHENER, G . D. (1996). Epilobium brunnescens hybrids in the British Isles. B.S.B.I. news 72: 65. Parmer, M. A. & Bratron, J. H., eds. (1995). A sample survey of the flora of Britain and Ireland. J.N.C.C., Peterborough. PRESTON, C. D. (1989). The spread of Epilobium ciliatum Raf. in the British Isles. Watsonia 17: 279-288. Raven, P. H. (1972). Evolution and endemism in New Zealand Epilobium, in VALENTINE, D. H., ed. Taxonomy, phytogeography and evolution, Symposium Manchester 1971. Academic Press, London. Raven, P. H. & Raven, T. E. (1976). The genus Epilobium (Onagraceae) in Australasia: a systematic and evolutionary study. New Zealand Department of Scientific and Industrial Research Bulletin 216, Christ- church, New Zealand. SCANNELL, M. J. P. & Synnotr, D. M. (1987). Census catalogue of the flora of Ireland. The Stationery Office, Dublin. Stace, C. A. (1975). Epilobium, in Stace, C. A., ed. Hybridization and the flora of the British Isles. Academic Press, London. Stace, C. A. (1976). Epilobium. B.S.B.I. news 14: 16-17. Stace, C. A. (1991). New Flora of the British Isles. Cambridge University Press, Cambridge. Wess, C. J., Sykes, W. R. & GaARNock-Jones, P. J. (1988). Naturalised pteridophytes, gymnosperms, dicotyledons. Flora of New Zealand 4: 884-903. Botany Division, Department of Scientific and Industrial Research, Christchurch, New Zealand. (Accepted May 1997) Watsonia 22: 61-68 (1998) 61 A reassessment of the hybrid Potamogeton X gessnacensis G. Fisch. (P. natans X P. polygonifolius, Potamogetonaceae) in Britain C. D. PRESTON I.T.E., Monks Wood, Abbots Ripton, Huntingdon, Cambs., PE17 2LS J.P. BAILEY Department of Botany, University of Leicester, Leicester, LEI 7RH and P. M. HOLLINGSWORTH Royal Botanic Garden, Inverleith Row, Edinburgh, EH3 5LR ABSTRACT Potamogeton X gessnacensis G. Fisch. (P. natans L. X P. polygonifolius Pourr.), hitherto known in Britain from Caernarvonshire (v.c. 49) and East Ross (v.c. 106), was discovered in 1996 in Shetland (v.c. 112). Cytological examination of plants from Caernarvonshire and Shetland shows that they are similar, with a chromosome number of 2n = c. 39. This is intermediate between P. natans (2n = 52) and P. polygonifolius (2n = 28), both counts previously reported from British material. Populations of P. < gessnacensis are variously intermediate in morphology between the putative parents, the Caernarvonshire plants being closer to P. natans than the others; they are highly sterile. The habitat of P. x gessnacensis at its British sites is described. P. natans has not been recorded from any of the sites, which suggests that it may have been eliminated by competition with the hybrid. Keyworps: Potamogeton coloratus, P. epihydrus, cytology. INTRODUCTION Potamogeton natans L. (Broad-leaved Pondweed) and P. polygonifolius Pourr. (Bog Pondweed) are the most frequent broad-leaved species of Potamogeton in the British Isles (Preston & Croft 1997). However, the hybrid between them appears to be very rare. Dandy (1975) reported “British plants which appear to be this hybrid”’ from two localities, Llyn Anafon, Caernatvonshire (v.c. 49) and Hill of Nigg, E. Ross (v.c. 106). Dandy’s cautious wording contrasts with his normally confident phraseology (other hybrids in the same account are described as “‘clearly intermediate’, ‘“obviously intermediate’, “manifestly intermediate” and “strikingly intermediate’ between the parental species) and suggests that he regarded the identification as in need of confirmation. Preston (1995a) also described P. X gessnacensis as a hybrid requiring further research, pointing out that plants in the two British populations differed in morphology, with those at Llyn Anafon approaching P. natans very closely. In August 1996 P.M.H. and C.D.P. discovered a further population of plants (at Loch of Gards, Shetland v.c. 112) which were morphologically intermediate between P. natans and P. polygonifo- © lius and appeared to be highly sterile. As P. natans and P. polygonifolius differ in chromosome number, fresh material was sent to J.P.B. for cytological study. His results supported the provisional identification of the Shetland plant as the hybrid P. X gessnacensis. The plant from Llyn Anafon was also examined cytologically and proved to be similar. The results of these cytological studies are presented in this paper, and discussed in relation to the morphology of the plants. 62 C. D. PRESTON, J. P. BAILEY AND P. M. HOLLINGSWORTH CYTOLOGICAL STUDIES Fresh material of putative Potamogeton X gessnacensis collected at Loch of Gards, Shetland (v.c. 112) on 1 August 1996 and Llyn Anafon, Caernarvonshire (v.c. 49) on 2 September 1996 was cultivated at the University of Leicester. Roots were pretreated in 8 hydroxyquinoline at 4°C for 8 hours, then fixed in fresh 3:1 ethyl alcohol:glacial acetic acid. The roots were then hydrolysed for 10 minutes at room temperature in 5N hydrochloric acid. The meristem was then dissected out in a drop of aceto-orcein using fine tungsten needles and tapped, flamed and squashed. Voucher specimens from both populations have been deposited in CGE. Chromosome counts of 2n = c. 39 were obtained from plants from Llyn Anafon (Fig. 1) and Loch of Gards (Fig. 2). The material is rather difficult cytologically. The chromosomes are rather numerous, very small (most are less than 1 wm long) and there is considerable variation in size between chromosomes within the complement. Another difficulty is that the chromosomes seem to be of two types (possibly the complements of the two parental taxa). One complement has a quadripartite appearance and resembles the textbook chromosome; it is clearly comprised of two chromatids, a centromere and four ends. The other sort of chromosome appears diffuse, without an obvious centromeric constriction and with only two ends apparent. The quadripartite chromosomes are easy to interpret even when adjacent, but this is not the case with the diffuse ones. In Fig. 1b (hollow arrow), the torpedo-shaped “‘body” has been interpreted as two adjacent chromosomes, though a case could be made for interpreting it as a single chromosome. However, other preparations from the same plant do not have a similarly shaped large chromosome. In Fig. 2b the V-shaped chromosome has been interpreted as a single chromosome with the chromatids splitting apart, rather than two small abutting chromosomes. The presence of a homologous chromosome is often useful in interpreting cases like this, but homologues need not be present in a hybrid (see the single strongly satellited chromosome arrowed in Figs 1a and 2a). A more certain count might be obtained by looking at the size and range of morphology of the chromosomes of the two parental taxa and checking the interpretation of the metaphase squashes in the light of this information. These interpretative difficulties are reflected in the fact that only approximate counts are presented here. Interpretative difficulties apart, the Llyn Anafon and Shetland karyotypes of P. X gessnacensis have a number of features in common. Both have a single very distinctive satellited chromosome (arrowed in Figs 1a and 2a), and the same mixture of ‘“‘quadripartite” and “diffuse” chromosomes. Potamogeton natans has a chromosome number of 2n = 52. This statement is based on eight & e*%, r e°* eb » a & * Py i of 3 ts . o? : v° . %@ sia 1b ‘ & FicureE 1. Cytological preparation of Potamogeton X gessnacensis from Llyn Anafon, Caernarvonshire (Fig. 1a) and the interpretation of this preparation as 2n = c. 39 (Fig. 1b). The solid arrows in Fig. 1a indicates a conspicuous satellited chromosome. The hollow arrow in Fig. 1b indicates an area which is difficult to interpret (see text). Scale: as Fig. 2. POTAMOGETON xX GESSNACENSIS 63 2b rae Ficure 2. Cytological preparation of Potamogeton X gessnacensis from Loch of Gards, Shetland (Fig. 2a) and the interpretation of this preparation as 2n = c. 39 (Fig. 2b). The solid arrow in Fig. 2a indicates a conspicuous satellited chromosome. The hollow arrow in Fig. 2b indicates an area which is difficult to interpret (see text). Scale bar = 2 um. separate counts, one (as 2n = c. 52) from English material (Hollingsworth et al. 1995; Hollings- worth et al. submitted). In this paper we have adopted, for convenience, the usual assumption that this represents the tetraploid level in Potamogeton. A count of 2n = 42 has also been reported for P. natans from two localities, in N. America (Stern 1961, as n = 21) and eastern Asia (Probatova & Sokolovskaya 1984, as 2n = c. 42). P. polygonifolius is a diploid species: 2n = 26 is reported for this species by Palmgren (1939, as n = 13), Fernandes (1950) and Ficini et al. (1980). However, the three English counts of this species all unambiguously gave a count of 2n = 28 (Hollingsworth 1995; Hollingsworth et al. submitted). In his summary of the chromosome numbers of Potamoge- ton, Les (1983) attributed the chromsome number 2n = 52 to P. polygonifolius on the basis of two (uncited) counts. We have only been able to trace a single apparent count of 2n = 52 for this species, by Takusagawa (1961). Although this count has been cited by abstracters (e.g. Moore 1973), a close reading of the text reveals that this count was never made, but was merely the chromosome number which Takusagawa would expect this species to have on the basis of the chromosome number of similar species. Judging by the chromosome counts cited above, one would expect the hybrid P. X gessnacensis to be a triploid with 2n = 39 or 2n = 40. The cytological results show that the plants from Llyn Anafon and Shetland are indeed triploid, with a number which is certainly close to the expected number. They are therefore almost certainly a hybrid between a diploid and a tetraploid species. The tetraploid species is clearly P. natans, on morphological grounds. Although the morphological evidence for the involvement of P. polygonifolius is less strong, this is almost certainly the diploid parent. The only other broad-leaved pondweeds in the British flora which are diploids are P. coloratus and P. epihydrus, neither of which are likely parents of a hybrid which grows in upland N. Wales and Shetland. RECOGNITION OF P. X GESSNACENSIS One of the features which distinguishes the hybrids of P. natans with the broad-leaved species P. gramineus (P. X sparganiifolius), P. lucens (P. X fluitans) and P. nodosus (P. X schreberi) is the presence of narrow, laminar submerged leaves intermediate between the phyllodes of P. natans and the laminar leaves of the other parent (Preston 1995a, b). Similar leaves might be expected to occur in the hybrid between P. natans and P. polygonifolius. However, the submerged leaves of the 64 C. D. PRESTON, J. P. BAILEY AND P. M. HOLLINGSWORTH British populations of P. X gessnacensis are much closer to the phyllodes of P. natans than they are to the laminar submerged leaves of P. polygonifolius. Plants of P. X gessnacensis from Llyn Anafon collected in May have phyllode-like submerged leaves which are sometimes expanded above the petiole into a very narrow lamina, and can occasionally resemble the narrowest submerged leaves of P. X sparganiifolius. These leaves have usually decayed by August or September, although they can sometimes be found on axillary shoots. The plants from Scotland have either phyllodes or phyllodes which are expanded at the tip into a lamina which resembles a rudimentary floating leaf. In the absence of leaves which are clearly intermediate between phyllodes and submerged leaves, the identification of P. X gessnacensis has to be based on features of the stipules and the floating leaves, which combine the characters of both parents (Table 1). Plants from Shetland are more easily recognised as intermediate than those from elsewhere, as they have short stipules which, when dry, are more or less translucent, lacking the very opaque, buff-coloured appearance of the P. natans stipule. When considered in conjunction with the phyllode-like submerged leaves and the presence on some leaves of at least a trace of the flexible junction between the petiole and lamina of the floating leaves, they clearly indicate that the plants are intermediate between the putative parents. The stipules of the other populations are closer to those of P. natans, but the floating leaves have longitudinal veins which are less markedly translucent than in that species. The flexible junction between the petiole and the lamina is also absent or poorly developed in these populations, but this character has to be used with caution as it is not always present in P. natans. The inflorescences of P. X gessnacensis are usually shorter than those of P. natans and even in the field it is possible to obtain some indication that it is a sterile hybrid. The perianth segments of plants at Llyn Anafon and Loch of Gards remain closed and the stigmas protrude through them. This behaviour is typical of sterile hybrids (Preston 1995a, p. 46) and contrasts with that of fertile species where the perianth segments open to expose anthers which shed copious amounts of pollen. At Llyn Anafon on 2 September 1996 the old inflorescences of the hybrid were rotting, and being replaced by newly developed inflorescences. Examination of the pollen of plants from Llyn Anafon has shown that it is sterile (G. C. S. Clarke, in litt. to J. E. Dandy, 10 July 1973, BM). Plants from Llyn Anafon sometimes produce easily detached fascicles of short phyllodes or phyllode-like leaves in the leaf axils: these were first noted by A. Fryer on a specimen collected by C. Bailey on 29 September 1884 (BM). Similar structures are occasionally produced by P. natans and may act as vegetative propagules. Material from Llyn Anafon is illustrated by Preston (1995a) and plants from Shetland are drawn as Fig. 3. BRITISH POPULATIONS OF P. X GESSNACENSIS Potamogeton X gessnacensis is now known from three sites in Britain. It has been collected at Llyn Anafon, Caernarvonshire, since 1884, the early specimens being originally named P. natans, P. polygonifolius or (in the case of plants collected by E. F. Cooper in 1890) P. natans xX polygonifolius. At Llyn Anafon P. X gessnacensis grows in water from 0-3 to at least 0-8 m deep around the edge of the lake. The stands of the hybrid are extensive, especially in relatively sheltered bays, typically occupying several square metres in area. They flower freely, but the inflorescences rot rather than set fruit. Llyn Anafon is a base-poor upland lake (altitude 500 m), although there may be slight base-enrichment from a small outcrop of dolerite and a larger outcrop of less basic andesite south of the lake. Its flora includes Callitriche hamulata, Isoetes lacustris, the aquatic variant of Juncus bulbosus, Littorella uniflora, Lobelia dortmanna, Menyanthes trifoliata, Myrio- phyllum alterniflorum, Potamogeton berchtoldii, Ranunculus omiophyllus, Sparganium angusti- folium, Subularia aquatica, Utricularia vulgaris sensu lato and Chara globularis var. virgata. Potamogeton polygonifolius occurs in runnels leading into the lake and in water up to 0-5 m deep near the edge, but P. natans is apparently absent. A second rare Potamogeton hybrid, P. X griffithii, also occurs at Llyn Anafon with one parent (P. alpinus) but not the other (P. praelongus). The record of P. X gessnacensis from a small stream on the Hill of Nigg, East Ross, is based on a single collection made by U. K. Duncan on 10 August 1970. In a letter to J. E. Dandy dated 5 November 1970 and now kept with the specimen in BM, she commented “‘I noted at the time that there was no typical natans to be seen in the vicinity on the Hill of Nigg. In fact I thought this must be 65 POTAMOGETON X GESSNACENSIS (ING) siueld saneio30A Om) Jo BuNsIsuod UsUIIDEds parIp 9[sUls & UO pase sUOTIRAIOSgQ , SS eee ee ee ee ee a ee JUdSaIg juosqy - juasqyv jUDSOIg SJINIJ POWIOJ-[JOM sonpold 0} Ajtoedea Zb-01 07-8 = 87-11 09-07 (wu) YyisueT SdDUdISOIOYUT SIQISIA STQISTA O[QISIA juosqe Ajares eulue] pue juasqy 2981} 10 JUDSqY 20e1} IO JUNSqY 20k1] IO JUSSgY ‘yuosoid AyjensQ gjoned usamjeq uoNoun! poinojoosiq (quoonjsue 10U (Moeu AIDA jnq eulWIe] UeY} Ajjensn jusonjsuen Jofed soumouios) - jl) Judonjsuen (aysr] 0) dn pjay voy) saved] anbedg aonbedo Jo onbedg jUSONsUeLI IT, YsolJ JO SUIOA [eUIPNIIsUOT SoAvg] SUTLOTLT onbedo + jusonysued} ‘ud013 YSIUMOIG onbedo onbedo jusonjsuey ‘uMOIg F ‘UMOIg IO UddIH + ‘Jynq Jo uMoig ‘JJNq IO UddIH (AIp uayM) Inojod 0S-O1 7r-07 88-ZE C8-0E OLI-OP (uur) qyduaT somndys Le OS, eS ee oe snyofiuosxjod «gq SPIeD jo yoo] ,93IN JO [IIH uojeuy UAT] supjou “gq ee ae ee SISUBIDUSSAB x “J ha a a Le EES ee Yee ee SISNAIVNSSAD X ‘d GIMG@AH UIAHL JO SNOILVINdOd HASILldd AHaHL AHL GNV SAITOFINODATOd ‘d ‘SNVLVN NOLZDOWVLOd AO SYALOVUAVHO ANOS AO NOSIUVdWNOO V ‘IT ATEAVL 66 C. D. PRESTON, J. P. B_ “LEY AND P. M. HOLLINGSWORTH Figure 3. Potamogeton X gessnacensis, drawn by L. T. Ellis from herbarium specimens collected at Loch of Gards, Shetland, by C.D.P. & P.M.H. on 1 August 1996 (Preston 96/145, 146, CGE). Scale bars = 3 cm. a form of that species. . . there were only four plants, all nearly the same, growing in a colony in the small streamlet at GR 28/828708. I had followed up the stream from the coast expecting to find the loch at approximately that locality, but to my disappointment it had completely disappeared, having become a “‘bog”’ full of moss, rushes etc., without even a pool of water left...” The third site, Loch of Gards, Shetland, is a shallow coastal lake. P. X gessnacensis grows in water 0-2—0-3 m deep, forming small patches over a sandy substrate around the edge of the lake, and larger stands by a wall which crosses the lake at the south end. The plants were flowering when first discovered on 1 August 1996 and still flowering when W. Scott revisited the site on 4 September 1996. There was no sign of fruit developing on either date. Loch of Gards is a relatively eutrophic lake which (judging by its situation) must receive some input of salt-spray. Few aquatic plants grow with the P. X gessnacensis, but the deeper water in the centre of the lake is dominated by Potamogeton pusillus and Ranunculus baudotii. Other species present include Littorella uniflora, POTAMOGETON xX GESSNACENSIS 67 Myriophyllum alterniflorum, Persicaria amphibia, Potamogeton filiformis and P. perfoliatus. Neither P. natans nor P. polygonifolius was recorded at Loch of Gards in 1996. It is remarkable that P. natans has not been recorded from any of the British P. x gessnacensis sites. At Llyn Anafon and Loch of Gards the habitat which appears to be suitable for this species is occupied by the hybrid, suggesting that if P. natans was formerly present it may have been out- competed by the hybrid. P. polygonifolius is present at Llyn Anafon in a different habitat. DISCUSSION The results of the cytological studies outlined above support the conclusion that the plants from Llyn Anafon and Loch of Gards are P. X gessnacensis. However, the evidence is not yet absolutely conclusive, and more detailed cytological study, or isozyme or other molecular studies, would be worthwhile. P. X gessnacensis is a difficult hybrid to detect in the field. In order to identify it one needs to recognise that it is intermediate between the parents, and resist the temptation to ascribe it to one parent or the other. It is unlikely that a population such as that at Llyn Anafon, which has a close resemblance to P. natans, could be identified on a single visit to a site, as the characteristic features of the submerged leaves are apparent only in the early season whereas the sterility of the plants can be established only later in the summer. The hybrid is more difficult to detect as a herbarium specimen, and the fact that J. E. Dandy was able to identify the plant from Llyn Anafon as P. x gessnacensis provides remarkable testimony to his knowledge of the genus. In view of these difficulties of identification, it seems likely that there are further populations of the hybrid in Britain that have hitherto escaped detection; the hybrid is also likely to occur in Ireland. The variation between the British populations of Potamogeton X gessnacensis is apparently paralleled by variation in the type locality, the Gessnach stream in Germany. In his initial description of the hybrid as P. X gessnacensis, Fischer (1907) recognised three varieties, one variety closer to P. natans, one closer to P. polygonifolius and one intermediate between the other two. SPECIMENS EXAMINED The above account is based on the following specimens of P. X gessnacensis. Determinations attributed to J. E. Dandy (J.E.D.) were made in 1973 unless otherwise stated. BRITISH ISLES WALES: Caernarvonshire, v.c. 49: Llyn Anafon, SH/69.69., 29 September 1884, C. Bailey, BM, det. J.E.D.; August 1890, E. F. Cooper, BM, det. J.E.D.;5 July 1905, H. W. Pugsley, BM, det. J.E.D.; 1 October 1906, J. F. Dutton, BM, det. J.E.D.; 30 July 1910, G. Goode, LTR, det. J.E.D., 1975; 4 August 1928, A. Wilson, YRK, det. C.D.P., 1997; 4 July 1946, R. Ross, BM, det. J.E.D.; 4 July 1984, M. Wade, UTLH; 28 May 1988, C. D. Preston & N. F. Stewart, Preston 88/35, 37, 38, BM, CGE, NMW; 27 August 1989, C.D.P. & N. F. Stewart, Preston 89/462, 464, 465, CGE, NMW; 2 September 1996, T. D. Dines & C.D.P., Preston 86/231, 232, 233, BM, CGE, NMW. SCOTLAND: East Ross, v.c. 106: N. side of Hill of Nigg, NH/828.708, 10 August 1970, U. K. Duncan, BM, det. J. E. D. Shetland, v.c. 112: Loch of Gards, Scat Ness, HU/38.09., 1 August 1996, P.M.H. & C.D.P., Preston 96/145, 146, 147, BM, CGE, E; 4 September 1996, W. Scott, Scott 3527, CGE, E, det. C.D.P., 1996. GERMANY In der Gessnach zwischen Odhof und Schaufling b. Deggendorf (Bayer. Wald), 17 August 1921, L. Oberneder, BM. ACKNOWLEDGMENTS We thank T. D. Dines and N. F. Stewart for help with fieldwork in Wales, and W. Scott for revisiting the Shetland site for P. X gessnacensis at our request. We are grateful to the B.S.B.I. 68 C. D. PRESTON, J. P. BAILEY AND P. M: HOLLINGSWORTH Bequest Fund for a contribution to the expenses of our visit to Shetland. Fig. 3 was drawn by L. T. Ellis and financed by a grant from the Bequest Fund for illustrations in Watsonia. REFERENCES Danby, J. E. (1975). Potamogeton L., in Stace, C. A., ed., Hybridization and the flora of the British Isles, pp. 444-459. Academic Press, London. FERNANDES, A. (1950). Sobre a cariologia de algumas plantas da Serro do Gerés. Agronomia lusitana 12: 551- 600. Ficini, G., GARBARI, F., GIORDANI, A. & ToMmEL, P. E. (1980). Numeri cromosomici per la flora Italiana. Informatore botanico italiano 12: 113-116. FIscHER, G. (1907). Die Bayerischen Potamogetonen und Zannichellien. Berichte der Bayerischen Botanischen Gesellschaft 11: 20-162. HoLiincswortH, P. M. (1995). Chromosome numbers, in Preston, C. D., Pondweeds of Great Britain and Treland, pp. 55-58. Botanical Society of the British Isles, London. HoLiiINcswortH, P. M., Preston, C. D. & GorNALL, R. J. (1995). Isozyme evidence for hybridization between Potamogeton natans and P. nodosus (Potamogetonaceae) in Britain. Botanical journal of the Linnean Society 117: 59-69. Ho.iincswortH, P. M., Preston, C. D. & GoRNALL, R. J. (submitted). Euploid and aneuploid evolution in Potamogeton (Potamogetonaceae). Submitted to Aquatic botany. Les, D. H. (1983). Taxonomic implications of aneuploidy and polyploidy in Potamogeton (Potamogetonaceae). Rhodora 85: 301-323. Moorg, R. J., ed. (1973). Index to plant chromosome numbers 1967-1971. Regnum vegetabile 90: 1-539. PALMGREN, O. (1939). Cytological studies in Potamogeton. Preliminary note. Botaniska notiser 1939: 246-248. PRESTON, C. D. (1995a). Pondweeds of Great Britain and Ireland. Botanical Society of the British Isles, London. PRESTON, C. D. (1995b). Potamogeton X schreberi G. Fisch. (P. natans L. X P. nodosus Poir.) in Dorset, new to the British Isles. Watsonia 20: 255-262. PresTon, C. D. & Crort, J. M. (1997). Aquatic plants in Britain and Ireland. Harley Books, Colchester. Propatova, N. S. & SoxoLovskayA, A. P. (1984). Chromosome numbers in the representatives of the families Alismataceae, Hydrocharitaceae, Hypericaceae, Juncaginaceae, Poaceae, Potamogetonaceae, Ruppia- ceae, Sparganiaceae, Zannichelliaceae, Zosteraceae from the Soviet Far East. Botanicheski zhurnal 69: 1700-1702. STERN, K. R. (1961). Chromosome numbers in nine taxa of Potamogeton. Bulletin of the Torrey Botanical Club 88: 411-414. TAKUSAGAWA, H. (1961). Cytological studies in the genus Potamogeton in Japan. Bulletin of the Shimane Agricultural College 9: 237-269. (Accepted April 1997) Watsonia 22: 69-82 (1998) 69 Potamogeton pectinatus L. X P. vaginatus Turcz. (P. X bottnicus Hagstr.), a newly identified hybrid in the British Isles C. D. PRESTON I.T.E., Monks Wood, Abbots Ripton, Huntingdon, Cambs., PE17 2LS P. M. HOLLINGSWORTH Royal Botanic Garden, Inverleith Row, Edinburgh, EH3 5LR and R. J. GORNALL Botany Department, University of Leicester, Leicester, LE] 7RH ABSTRACT A hybrid Potamogeton (Potamogetonaceae) which grows in shallow, rapidly flowing water in the River Till (Cheviot) and River Tweed (Berwickshire (v.c. 81) and Cheviot (v.c. 68)) is apparently referable to P. x bottnicus Hagstr. (P. pectinatus L. x P. vaginatus Turcz.). Both morphological and isozyme characters are consistent with this hybrid combination rather than with the previous identification of these plants as P. x suecicus K. Richt. (P. pectinatus x P. filiformis Pers.). British plants of P. x bottnicus and European plants of P. vaginatus are described and the differences between them, P. pectinatus and P. X suecicus outlined. Possible explanations for the presence of P. X bottnicus in Britain in the absence of P. vaginatus are discussed. The name P. X bottnicus is lectotypified by a syntype referable to P. pectinatus x P. vaginatus. Examination of type material of P. X meinshausenii Juz., described from the vicinity of St Petersburg, indicates that it is not a hybrid between P. pectinatus and P. vaginatus as hitherto supposed. KeEyworps: Potamogetonaceae, Potamogeton filiformis, P. X suecicus, P. X meinshausenii. INTRODUCTION Two species in Potamogeton subgenus Coleogeton Rchb. are found in the British Isles, P. filiformis Pers. (Slender-leaved Pondweed) and P. pectinatus L. (Fennel Pondweed). The hybrid between them, P. X suecicus K. Richt., is one of the more difficult Potamogeton hybrids to identify, partly because of the extreme variability of the commoner parent, P. pectinatus, and partly because the crucial character which separates this parent from the hybrid, the structure of the leaf sheaths, can only be ascertained by dissection under the microscope. The first correct records of P. X suecicus in the British Isles were published by Dandy & Taylor (1940). The hybrid has subsequently been discovered in scattered localities in Scotland and Ireland, where both the widespread P. pectinatus and the more northerly P. filiformis occur (Preston & Croft 1997). More significantly, P. x suecicus was reported by Dandy & Taylor (1946) from the River Tweed in Berwickshire (v.c. 81) and Cheviot (v.c. 68), and the River Wharfe and River Ure in Yorkshire (v.c. 64 and v.c. 65); a similar plant has subsequently been found in a tributary of the Tweed, the River Till (Dandy 1975; Holmes & Whitton 1975a, b; Swan 1993). At the time of Dandy & Taylor’s paper these localities all lay south of the extant sites for P. filiformis in Britain. P. filiformis has since been found in Rayburn Lake, S. . Northumberland (v.c. 67), south of the Tweed and the Till, but the Yorkshire sites are still some 150 km south of the nearest known P. filiformis population. We have recently re-investigated P. X suecicus in Britain, examining the morphology of populations and using isozyme electrophoresis to investigate the variation in both the hybrid and its putative parents (Hollingsworth et al. 1996a, b). Most populations hitherto identified as P. x 70 C. D. PRESTON, P. M. HOLLINGSWORTH AND R. J. GORNALL suecicus, and a number of new populations discovered during the course of this study, have proved to be intermediate morphologically between P. filiformis and P. pectinatus, and the results of the isozyme analysis offered strong support to the hypothesis that they represent the hybrid between these two species. This applies both to plants in sites in Scotland, where the hybrid often grows in proximity to both parents, and to the outlying populations in the River Wharfe and River Ure. Hollingsworth et al. (1996a) concluded that the evidence that the populations studied are the hybrid between P. filiformis and P. pectinatus is “‘virtually conclusive’. An updated account of the distribution of P. X suecicus in the British Isles is in preparation. Although the identity of most populations of P. < suecicus was confirmed by our recent study, the plants in the River Till and River Tweed were a conspicuous exception. We were unable to identify them as P. X suecicus either on morphological grounds or from the isozyme evidence. Both morphology and isozyme evidence is, however, consistent with the hypothesis that these plants are P. X bottnicus Hagstr., the hybrid between P. pectinatus and the third European member of subgenus Coleogeton, P. vaginatus Turcz. This is a surprising conclusion, as in Europe P. vaginatus is confined to Norway, Sweden and Finland (Elven & Johansen 1984); the nearest population lies some 1500 km from the sites in the River Tweed and River Till. The species is also found in central Asia and is widespread in North America (Hultén & Fries 1986). The hybrid is described below, and the evidence for its identity outlined. POTAMOGETON X BOTTNICUS IN BRITAIN DESCRIPTION The following description of P. x bottnicus is based on fresh material and herbarium specimens collected from the British sites. Plants forming large and vigorous clumps. Rhizomes to at least 0-45 m long, 1-5-6-5 mm in diameter. Stems up to 1-5 m long, 1-7-4-3 mm in diameter, terete, with frequent branches lying more or less parallel to the main stem; nodal glands absent. Scales 10-64 mm long, present on lowest (O—)1-3 nodes of the stem, clasping the stem throughout their length or with the distal part free, incurved and leaf-like, or bearing rudimentary leaves. Submerged leaves linear, mid to dark green, sometimes tinged with brown when growing near the surface of the water, leaves at the first 4 nodes above the basal scales 25-206 xX (1:2—)2-0-3-5 mm, 11-90 times as long as wide, 0-8-3-7 times as long as the sheath, stiff, markedly canaliculate, acute; leaves towards the apex of mature stems (72—)110— 200(—250) x 0-8-3-1 mm, (45—)60—110(—140) times as long as wide, (2-6—)3-0—S-7 times as long as the sheath, canaliculate, more or less truncate, rounded or obtuse at the apex, sometimes slightly mucronate and often slightly asymmetrical, sometimes more or less acute on flowering stems, entire and plane at the margin, the midrib bordered on each side by 1-2 inconspicuous lateral veins and several air channels. Floating leaves absent. Leaf sheaths green with hyaline margins, the margins sometimes with a brownish tinge, open and usually convolute unless forced apart by branches arising at the node, 25-78 x 2-1—5-5 mm at the first 4 nodes above the basal scales, 23-70 mm long towards the apex of mature vegetative stems, only 17-28 mm long towards the apex of flowering stems; ligules 6-17 mm, hyaline, rounded or truncate at the apex. Turions absent. Inflorescences 10-26 x 4-5-5 mm; peduncles 36-90 Xx 0-6—-0-7 mm, pale pink, terete, flexuous. Flowers 5-10, in 3-5 groups of 1-2; anthers small, hidden by the tepals, not filled and not dehiscing readily, the pollen misshapen; carpels 4, the stigmas sessile. Fruits not seen, and almost certainly do not develop. Vegetative reproduction by small plantlets which develop on short axillary stolons; perhaps also by tubers which are formed on the rhizomes of both parents but these not yet seen on the hybrid. COMPARISON OF P. X BOTTNICUS WITH P. PECTINATUS AND P. X SUECICUS The salient characters distinguishing the three European species in subgenus Coleogeton and the hybrids P. x bottnicus and P. X suecicus are set out in Table 1. British and Irish botanists are most likely to overlook P. X bottnicus as a form of the variable P. pectinatus. Fortunately, both P. pectinatus and P. X bottnicus grow together in the River Tweed on the S. side of St Thomas’s Island, Norham Mains, and in July 1995 the differences between the taxa were clearly apparent in this mixed stand. The clumps of P. X bottnicus reached the surface, where the leaves were brownish green, and scarcely moved in the current. The leaves had blunt apices and long, broad sheaths. The POTAMOGETON PECTINATUS X P. VAGINATUS 71 clumps were flowering rather sparingly, with approximately 160 inflorescences per square metre, and the anthers were hidden by the tepals and did not appear to be releasing pollen on to the water surface. P. pectinatus formed bright green clumps which at that time did not reach the surface of the water but were waving in the current. The plants were richly branched, and it was difficult to distinguish a single main stem. The leaves were narrower and more finely tapered towards the apex than those of P. x bottnicus, and their sheaths were shorter and more slender. Flowers were more frequent than on the clumps of P. X bottnicus, with approximately 600 inflorescences per square metre of water surface; the anthers exceeded the tepals, were clearly full, and dehisced on the surface of the water to release pollen which drifted downstream. Pollen from both taxa in this stand was subsequently examined microscopically, and the well-formed pollen of P. pectinatus contrasted with the misshapen pollen of P. x bottnicus (Table 2). Measurements based on stems of P. X bottnicus and P. pectinatus collected at random from the mixed stand in the R. Tweed and the population of P. x bottnicus in the River Till are presented in Table 3 and Fig. 1. These clearly illustrate the vegetative differences between the two taxa at these sites. One complicating factor in comparing the two taxa is the ontogenetic variation which is shown by most members of subgenus Coleogeton: lower leaves (especially on short pioneer shoots) tend to be broader and more obtuse than upper leaves, and leaves on vegetative stems tend to be broader than those on flowering stems. The leaves at the apex of long flowering shoots of P. X bottnicus are, therefore, more like those of P. pectinatus than other leaves. P. pectinatus is a very variable species and some populations, such as those which were formerly segregated as P. flabellatus Bab., may be indistinguishable vegetatively from P. x bottnicus. However, the sessile stigmas of the hybrid are crucial in distinguishing it from all forms of P. pectinatus which may resemble it vegetatively. In vegetative characters P. x bottnicus is usually closer to P. X suecicus than P. pectinatus, and in particular it resembles the robust forms of that hybrid in the R. Wharfe and R. Ure. Characteristic specimens of P. X suecicus from these rivers are illustrated by Dandy & Taylor (1946) and Preston (1995). P. X bottnicus and P. X suecicus are both sterile hybrids with obtuse leaves and sessile stigmas. The large lower leaf sheaths of P. x bottnicus might be thought to be derived from the similar sheaths of P. vaginatus (which they closely resemble) and thus provide a character to distinguish it from P. X suecicus, but some pioneer shoots of the latter have surprisingly large sheaths, as do some plants of P. pectinatus. The crucial distinction between P. X bottnicus and P. X suecicus lies in the fact that the leaf sheaths of the former are open whereas at least some of the sheaths of P. X suecicus are closed and tubular at the base. Data for selected populations are provided in Table 4. In interpreting this table it should be remembered that a closed sheath is an unambiguous character, whereas a sheath may be recorded as open because it actually is open, or because it is closed for an indetectably short distance above the base, or because it is a closed sheath which has split. Nevertheless, the data in Table 4 (which are based on the dissection of young and apparently intact sheaths) demonstrate that in some P. X suecicus populations all the sheaths are tubular, whereas others consist of plants with a mixture of open and tubular sheaths. The following key to the British and Irish taxa in subgenus Coleogeton can be used in place of that presented by Preston (1995, p. 133). MPa MER IaSTOP CIA CO UAC WANE) ase. cs cele aden eae. doko hisueube sha Ribmleeede oem eeeeede Aloe wil al 2 emsemcrwer ail sheaths tubular at the DASE ....5... 0.0... scscbecsesecacsecsdueccceetcoacmastnlesccaccacdeweess 3 2a. Mature leaves on the vegetative stems usually acute to finely acuminate at the apex; stigmas borne on a distinct style c. 0-2 mm long; pollen well-formed; fruits 3-3-4-7 mm_ P. pectinatus 2b. Mature leaves on the vegetative stems more or less truncate, rounded or obtuse at the apex; stigmas sessile; pollen misshapen; fruits not developing .........................0005 P. X bottnicus 3a. Stems branched at base, otherwise unbranched or very sparingly branched; all sheaths tubular at the base; stigmas sessile; pollen well-formed; fruits 2-2—2-8(-3-2) mm .......... P. filiformis 3b. Stems usually sparingly or richly branched above the base; all sheaths tubular at the base or some tubular and others open and convolute; stigmas sessile or borne on a distinct style; pollen MMSSADEM-VATiits\ NOL GCVELOPIRE) so. buis2.s0c 22. enabupintane gad sb abilens «he shebwbabaleneaee ys P. X suecicus C. D. PRESTON, P. M. HOLLINGSWORTH AND R. J. GORNALL 72 Se 8-€-9-7 = L-v-€-€ as i : (Z-€-)8-7-C-Z (wu) WMI a[N104 aT91¢ a[N404 STLBEHIS a[n41e4 uaT[Od g]Issas g]ISsos poyxeis PoyY[eIs 10 o]Issag O]ISSag seuss p-S-0 LI-9 SIs v7-L Gies (urur) sopnsry] oJeUIUINOe SOABOT gynoe O} 9sSNIGO 9ynoe O} 9JeVoOUNI oyeUIWINOe O} INnIV 0} gINoeQGNnsS gynoeke O} 9SNIQO youelg Jo xody SOALO] YOURIQ WO SULIOJJIP SOACD] JUdSIIg JUISoIg jussoid sowMouoS juasoid sowmMaulos juasqy WI9}s peolq ‘1104S uodo owios ‘1ejnqny uodo usdoO uodo 9WIOS IO Ie]NQnN], Ieynqny, syjeoys jeaT SNIDUISDA ‘q sno1Uyjog x ‘d snyouygad *g SN1IANS x ‘df smmsofyyf *g VIAVNIGNVOS WOU SALVNIOVA ‘d GNV ‘SHTISI HSILIYG AHL WOU (SALVNIDVA X SQLVNILOUd ‘d) SQOINLLOY X ‘d ANV (SALWNILOAd x SINYOAITIM ‘d) SQOIDANS X ‘d ‘SNLWNILOAd ‘d ‘SINYOATTIA NOLZDOWVLOd AO SHALOVUVHO AWOS JO NOSIUVdNOO ‘I ATAVL POTAMOGETON PECTINATUS Xx P. VAGINATUS 73 TABLE 2. POLLEN STAINABILITY OF POTAMOGETON PECTINATUS AND P. x BOTTNICUS FROM A MIXED STAND IN THE RIVER TWEED, AND P. VAGINATUS FROM FINLAND Taxon Locality Date Pollen stainability (%) P. pectinatus St Thomas’s Island, Norham Mains 1995 86-6 P. X bottnicus St Thomas’s Island, Norham Mains 1995 0-8 P. vaginatus Kuljunniemi, Saloinen, Raahe 1987 93-9 P. vaginatus Martinlahti, Piehinki, Raahe 1992 95-9 Stainability based on at least 300 grains from at least three anthers stained with safranin in glycerol. Based on material of P. pectinatus and P. X bottnicus collected by the authors and herbarium specimens of P. vaginatus collected by J. Sarkké (OULU and herb. J.Sarkka). —, Ol ep) (=) ©) (-) Sheath length (mm) (Se) &) 20 10 0.0 0.5 1.0 1.5 2.0 2.5 3.0 35 Leat width (mm) Figure 1. Sheath length (mm) plotted against leaf width (mm) for Potamogeton X bottnicus and P. pectinatus, based on measurements at the lowest four leaf-bearing nodes of randomly sampled fresh vegetative stems collected on 18-19 July 1995. Samples of P. X bottnicus collected from the River Tweed at Norham Mains (solid squares; n = 40) and the River Till below Twizel Bridge, NT/883.433 (solid triangles; n = 46). Samples of P. pectinatus (open circles; n = 43) collected from the same site as the River Tweed P. X bottnicus. COMPARISON OF P. X BOTTNICUS WITH P. VAGINATUS In Europe Potamogeton vaginatus is restricted to Norway, Sweden and Finland. We are not aware of a detailed description of the European plant in English, and therefore include one here. The following description has been drawn up primarily from Finnish material in OULU and in the private | herbarium of J. Sarkka, supplemented by specimens from Finland and Sweden in BM, CGE and LTR. As it is based entirely on dried specimens, it is likely to under-estimate the range of variation of the species, especially in features such as stem length, and it lacks characters which cannot be deduced from pressed plants. The illustration of P. vaginatus in Fryer & Bennett (1915) is based on C. D. PRESTON, P. M. HOLLINGSWORTH AND R. J. GORNALL 74 ‘OPW JOU 919M YOIYM SJUIWIIINSeOUI Sd}eoIpUI Ysep VY ‘(4) ZI pue (A) €€-02 “(y-T SPPOU) ZI-L ‘ozIs e[dureg “susois saneiosea soddn oy uo (10-0>d) yrpim Jeo] pure (¢Q'9>d) yISueq Jeo] ‘(SQ-0>d) yISUE] YIeays 105 pue “(¢0-9>d) p 2pou Je YISUT Jeo] 10} saajosway suotefndod snoiuyjoq x ‘_q dy} UI9M19q S2dUdIATFIP JUROYIUSIS ATTeoNSNEIS OsTe oIe S10YL, *(100-O>d ‘1891 A Aouity\-uUey) [ apou ye yIsuI] Jeo] 10} esoy} Jdaoxe JuRoYrUsIs AT[eoNsHeIs a1e suoNe|ndod snoiu0g x ‘g ay) pue snwuYoad ‘g UZEMI12q S2DUSIATJIP SUT, ‘swiais (4) SULIOMOY pure (A) 2ane1980< Jo sopou Joddn oy} pur ‘spremdn ([) aseq oy} WoIy poJoquinu ‘sopou BuLIedDq-Jed] INO] 1S9MO] DY} IO] poptaoid ore sonqeA “Ep €88/,LN ‘espug [OZIM] MOTO [ILL JOArYy Pure “C6 €Z6/LN ‘Sule WeYION 18 poaMy JoaAry oY] WOT] C66 AINE 6[-B] UO pajdaT[09 surais pofdures AyWOpueI UO paseq s}UdUIOINsRapy = i = eT (GAL GPA (AL = = 16 C6 IS vL = 5a 87 87 SC 97 WUINUWITXe YA] = ae 60 60 60 #£46:0 5 fe €S €S LE Of = = O07 4 8I fill ues] Set te 920s S$: S00 = = €C O07 LI i = = cl De iat OL WUNWTUTTAy (poom], “Y) Snjvuijzad ‘g CulepssG 4 esse (8 eGse “gee = CCT i. — : gs ty hits fi LFA, Po ak Eyer E32 WSIS A os) Ld Vy } 5 4 fy sd r Ws aa aews i a Wea OU S A 7 . iy Aa Ti a! buae i ‘ ‘iS fetes, ee my cour 5 ate ‘i ; Oe tat Wee is a rivet ; his aay S203 ; i iy ret Mahe 7 ey at “ ; * 7 i "7 ° al! . se b 4 Eo. j ae ae nD 17 , i y eT EB bh f ee ES Ty Dee rivera og ae P , ‘ d re a hey servsecs i r 2 $ tolenie.( ren hy Sei my? ie a eo ae ie f 459 pe hayes i Fi 5 Seed PRD ER PEE nee ae Sse ‘ , ‘ Ya eee ‘ . ‘i “pon ty yi at, a 4 bib 7 Bie tie: Rs ROIS fetter bg wt j ME) A f % j i ‘ e x ¥ - ae i" ; j in 2 ¥ ‘ 4 ! y $ ; he tet P » aS ‘ , im *y . ; H si } Ah ; met Gad 74 4 uJ i ); ‘ ve , { ; Ai: A + as T iy \ itera ub gabveryh inly in. Beh WOES REY i AAA pidbiny®! ‘ele i a th ~ ’ ef P : Pa } ; } } : opr te aes ‘ My 4" : i >, v: ‘. ‘ oy kt vat VAT ES CY ha TRS EVa ERE Beats tide Hf ’ sax 1g Thea ys ee AGRE ee Seri eyetegi' 4 Acs aan 7 , j ‘ wn eee " ™ ‘ et 4 Hi f ‘ ’ A +), "7 » PE ; Ss as. ; yy ie ine 2 (8 v d ‘ , y bi rh SORE aR” sae ‘ { : ue hae } R . 4 b yt Ot BN f - ” in Watsonia 22: 109-116 (1998) 109 Notes NORTHERN LIMITS ATTAINED BY NATIVE BRITISH PLANTS IN NORTH PEARY LAND, GREENLAND In June 1995 I was able to visit North Peary Land, Greenland with an Arcturus Expedition led by Robert Burton. The botanists in the party were Dr Jean Balfour, Dr Hugh Lang, Dr Fritz and Mrs Elizabeth Schwarzenbach and myself. Our base camp was at the head of Frigg Fjord at 83° 11’ N latitude some 50 km to the south of Kap Morris Jesup, the most northerly point of land in the world. Cape Columbia, the most northerly point of the Canadian arctic, on Ellesmere Island is at latitude 83° 08’ N. During our two week stay we recorded some 80 species of vascular plants. Twenty five of the species seen also occur in Britain and Ireland and of these, four were found at a new northern limit (Table 1). Apart from Kobresia simpliciuscula which also occurs in Upper Teesdale, these four species are confined to the Scottish Highlands. Christian Bay, a botanist with the Greenland Botanical Survey, had visited Frigg Fjord in 1985 and was impressed with the relative richness of the flora at this high latitude. He considered it to be a high arctic oasis. He also visited many other Peary Land localities including Kap Morris Jesup (Fredskild et al. 1986, 1987). British Floras, e.g. Clapham, Tutin & Warburg (1952), Sell & Murrell (1996), continue to quote the same north latitude figures for several of the species found in North Peary Land. I have therefore TABLE 1. THE NORTHERN LIMITS ATTAINED BY THE 25 NATIVE BRITISH PLANTS IN NORTH PEARY LAND, GREENLAND Species Site name Northern limit Alopecurus borealis Kap Morris Jesup 83°39'N Cardamine pratensis Brainard Sund 82°58'N *Carex atrofusca Frigg Fjord 83°12'N Carex maritima Frigg Fjord 83°16'N Carex rupestris Brainard Sund 82°58'N * Carex saxatilis Frigg Fjord 83°16'N Cerastium arcticum s.1. Kap Morris Jesup 83°39'N Cystopteris fragilis s.1. Frigg Fjord 83°12'N Equisetum arvense Frigg Fjord 83°12'N Equisetum variegatum Frigg Fjord 83°16'N Juncus biglumis Kap Morris Jesup 83°39'N *Juncus castaneus Frigg Fjord 83°12'N Juncus triglumis Frigg Fjord 83°16'N *Kobresia simpliciuscula Frigg Fjord 83°12'’N Koenigia islandica Frigg Fjord 83°12'N Minuartia rubella Kap Morris Jesup 83°39'N Oxyria digyna Constable Bugt 83°34'N Poa glauca Frigg Fjord Soul Persicaria vivipara Constable Bugt 83°34'N Sagina nivalis Frigg Fjord 83°12'N Saxifraga cernua Kap Morris Jesup 83°39'N Saxifraga cespitosa Kap Morris Jesup 83°39'N Saxifraga nivalis Kap Morris Jesup 83°39'N Saxifraga oppositifolia Kap Morris Jesup 83°39'N Silene acaulis (leg. O. Bennike 1984) Nansen Land 83°09'N The nomenclature of the species follows Stace (1991). *New northern limit 1995. 110 NOTES documented updated information on these hardy members of the British flora found at the most northern botanical localities on earth. The information given in Table 1 is based on Bay (1992), Bay (pers. comm. 1997) and the 1995 field work at Frigg Fjord. Not surprisingly 21 of the 25 species listed are members of the arctic-alpine or arctic-subarctic element of the British flora (Matthews 1955). Equisetum variegatum is a representative of the northern montane element and Cardamine pratensis, Cystopteris fragilis and Equisetum arvense are widespread (Birks 1973). Seventeen of the species are designated rare or scarce (Perring & Farrell 1983; Stewart et al. 1994). In addition to the three widespread species only Juncus triglumis, Oxyria digyna, Persicaria vivipara, Saxifraga oppositifolia and Silene acaulis are relatively common and widespread in the Scottish Highlands but much scarcer elsewhere in the British Isles being rare to very rare in Ireland from which Juncus triglumis is absent (Webb 1977). All the species are well distributed throughout most of the Arctic with Juncus triglumis being represented by the closely related Juncus albescens in the Canadian arctic (Polunin 1959). They are all common and widespread in Greenland with the following exceptions: Alopecurus borealis is absent from the southern half and Cardamine pratensis, Carex atrofusca, C. rupestris, C. saxatilis, Juncus castaneus and Kobresia simpliciuscula show disjunct patterns of distribution. This may be related to areas which have undergone little or no glacial erosion during the Weichselian (Devensian) glaciation (Bay 1992). The reported increase in the northward range of these species is due to fieldwork conducted in new areas rather than new colonisation from possible climatic amelioration. The four species seen at a new northern limit in 1995 gave the appearance of being long established. The favourable south facing Carex stans mires irrigated by snow beds and melt from the underlying permafrost were in stark contrast to the areas of dry stony high arctic desert virtually devoid of plant life. ACKNOWLEDGMENTS I am grateful to Christian Bay for checking the plant collections and help with the latitudinal data. I also thank Bent Fredskild, Director of the Greenland Botanical Survey, for copies of the survey reports. Finally I would like to thank Robert Burton, Neville and Kathleen Cartwright of Arcturus Expeditions for making the 1995 field work possible. REFERENCES Bay, C. (1992). A phytogeographical study of the vascular plants of northern Greenland — north of 74° northern latitude. Meddelelser om Grgnland. Bioscience 36. Birks, H. J. B. (1973). Past and present vegetation of the Isle of Skye. Cambridge University Press, Cambridge. CLAPHAM, A. R., Turin, T. G. & WarBurG, E. F. eds. (1952) Flora of the British Isles, 1st ed. Cambridge University Press, Cambridge. FREDSKILD, B., Bay, C., Hort, S. & NIELSEN, B. (1986). Gronlands botaniske undersogelse 1985, pp. 26-33. Botaniske Museum, Copenhagen. FREDSKILD, B., BAY, C. & PETERSEN, F. R. (1987). Gronlands botaniske undersogelse 1986, pp. 27-33. Botaniske ' Museum, Copenhagen. MartrHews, J. R. (1955). Origin and distribution of the British flora. Hutchinson’s University Library, London. PERRING, F. H. & FARRELL, L. (1983). British red data book: I vascular plants, 2nd ed. R.S.N.C., Lincoln. PoLunin, N. (1959). Circumpolar arctic flora. Oxford University Press, Oxford. SELL, P. D. & MurRELL, G. eds. (1996). Flora of Great Britain and Ireland, Vol. 5. Cambridge University Press, Cambridge. STAcE, C. A. (1991). New flora of the British Isles. Cambridge University Press, Cambridge. STEWART, A., PEARMAN, D. & Preston, C. D. (1994). Scarce plants in Britain. J.N.C.C., Peterborough. Wess, D. A. (1977). An Irish flora, 6th ed. Dundalgan Press, Dundalk. R. W. M. CorNER Hawthorn Hill, 36 Wordsworth Street, Penrith, Cumbria, CAl1 7QZ NOTES 111 POPULATION SIZES OF GENTIANELLA ULIGINOSA (WILLD.) BOERNER, DUNE GENTIAN, ON COLONSAY (V.C. 102) IN 1996 Gentianella uliginosa (Willd.) Boerner, Dune Gentian, has been recorded from five sites in South Wales (Lousley 1950; Abell 1954; Pritchard 1959; Kay 1972; Ellis 1983), three sites on Colonsay (Rose 1998) and two old sites in England (Rich 1996). In July and August 1996, a detailed survey of two of the three sites on Colonsay was undertaken as part of the Scottish Rare Plant Project (Lusby 1992) of the Royal Botanic Garden, Edinburgh. At Balnahard Dunes seven populations were found, with a maximum separation of c. 300 m. 1709 plants were counted in the largest population by R. L. Gulliver and P. Lusby using a grid. Each plant had a dried chick pea placed beside it to avoid counting it twice. Individuals were assigned to G. uliginosa rather than G. amarella if the pedicel length was well above 50% of the plant height (Pritchard 1959; Stace 1991; cf. Pritchard & Tutin 1972). Most G. uliginosa and some G. amarella and G. campestris plants were single flowered and less than 8 cm tall. Occasionally specimens of G. uliginosa with only a single leaf at the base of the pedicel were encountered. (In one case a plant had a single leaf at the base of one of its pedicels and two leaves at the base of the other two, thus confirming the status of the single structures as leaves.) The size of the other six populations was estimated, with values ranging between 200 and 1000. The total estimate for the site was 4509. In 1994 the sizes of the two largest populations in South Wales, where numbers are known to vary greatly from year-to-year (Q. O. N. Kay, pers. comm., 1997) were 4000-8000 at Oxwich and 600- 1000 at Whiteford (Kay & John 1995). Colonsay therefore contains one of the largest known populations of G. uliginosa in the British Isles. The number of plants of all three species of Gentianella with fully developed corollas (in terms of length) and/or with capsules was recorded on 24 July, 9 August and 23 August 1996 from a fixed 2 x 2 m quadrat (Table 1). The values in Table 1 can increase in time as small flower buds grow to their full length; or decrease due to grazing. On 24 July none of the G. uliginosa plants recorded were in fruit, by 9 August half were completely in fruit and a further nine had capsules plus either closed or open corollas. By 23 August the majority of plants were in fruit. G. campestris was the latest flowering of the three species, with G. amarella occupying an intermediate position, though the number of plants present was small. At Balnahard in 1996 sheep, cattle and rabbits were present and grazing levels were high in July and August. This seemed to act “preferentially” on the taller G. amarella thereby reducing the TABLE 1. THE NUMBER OF PLANTS WITH FULLY DEVELOPED COROLLAS AND/OR CAPSULES OF GENTIANELLA ULIGINOSA IN A FIXED 2 X 2M QUADRAT IN JULY AND AUGUST 1996 AT LEAC BHUIDHE, BALNAHARD DUNES, COLONSAY, V.C. 102, TOGETHER WITH NUMBERS OF GENTIANELLA AMARELLA AND GENTIANELLA CAMPESTRIS (SEE ALSO TEXT) 24 July 9 August 23 August Gentianella uliginosa Corolla(s) closed (fully expanded but not open), no capsules 22 8 0 Corolla(s) open, no capsules 70 0 4 Corolla(s) closed, capsules also present 0 8 0 Corolla(s) open, capsules also present 0 1 0 Total number of plants, flowering or about to flower Vip 17 5 Capsules (only) 0 iy 28 Total number of Gentianella uliginosa plants with fully expanded corolla(s) jis 34 38) Gentianella amarella Corolla(s) present + 4 5 Gentianella campestris Corolla(s) present + 25 62 Note + indicates present, no count undertaken. 112 NOTES possibility of introgression between the two species, which is reported to be affecting G. uliginosa populations in South Wales (Pritchard 1959, 1972). On the other hand Kay & John (1995) report more or less pure stands of G. uliginosa from South Wales with little or no signs of introgression, even when surrounded by G. amarella. Grazed examples of G. uliginosa were noted on all three visits. In 1996 the locations of the Balnahard populations (slopes, mini-plateaux and dry hollows) were all very dry. They frequently contained small gaps in the dune turf in which annuals could establish. In South Wales the plants usually grow in dune slacks with Salix repens present, though the water table is often well below the surface all the year round (Q. O.N. Kay, pers. comm., 1997). Lousley (1950) reported some plants on dry dune grassland at Oxwich Burrows. On Colonsay no G. uliginosa was found at the second site, Traigh nam Barc, where it was reported in 1981 by Rose (1998); though G. amarella, G. campestris and Coeloglossum viride which are associated with it at Balnahard Dunes were present. Kiloran Bay (Traigh Ban), the third site with a 1981 record by Rose (1998), has been examined repeatedly between 1991 and 1996. Neither G. uliginosa, G. amarella nor Coeloglossum viride have been located. However G. campestris has recently (1997) been found (A. Skrimshire, pers. comm., 1997). G. uliginosa is a British Red data book species which has been included on the list of species to receive special conservation attention as part of the United Kingdom’s contribution to the United Nations Convention on Biological Diversity. More information on year-to-year fluctuations in population numbers, the precise habitat/management requirements of the species, and on the extent of hybridization with G. amarella are urgently needed, especially as Colonsay represents the only known station for this rare and elusive plant in Scotland, where it is at the most north west edge of its global distribution (Hulten & Fries 1986). ACKNOWLEDGMENTS I would like to thank all who have helped in this study, particularly Mr Douglas Gilbert, Mr Graham Grant, Mr Nigel Grant, Mrs Mavis Gulliver, Mr David Hobhouse, Mr Andy Jones, Dr Quentin Kay, Mr Phil Lusby of the Royal Botanic Garden, Edinburgh, Scottish Rare Plant Project, Mr Alasdair McNeill, Dr Tim Rich, Dr Francis Rose, Miss Angela Skrimshire and Professor C. A. Stace. REFERENCES ABELL, R. B. (1954). Plant records — Gentianella uliginosa. Proceedings of the Botanical Society of the British Isles 1: 57. EL.is, R. G. (1983). Flowering plants of Wales. National Museum of Wales, Cardiff. HUuLTEN, E. & Fries, M. (1986). Atlas of North European vascular plants north of the Tropic of Cancer 2. Koeltz Scientific Books, Koenigstein. Kay, Q. O. N. (1972). The dune gentian in the Gower peninsula. Nature in Wales 13: 81-85. Kay, Q. O. N. & Joun, R. F. (1995). The conservation of scarce and declining plants in lowland Wales: population genetics, demographic ecology and recommendations for future conservation in 32 species of lowland grassland and related habitats (Countryside Council for Wales Science Report No. 110). Countryside Council for Wales, Bangor. (Gentianella uliginosa, pp. 95-103). LousLey, J. E. (1950). The habitats and distribution of Gentianella uliginosa Willd. Watsonia 1: 279-282. Lussy, P. (1992). Conservation of rare plants in Scotland. Botanical Society of Scotland news 58: 2-4. PRITCHARD, N. M. (1959). Gentianella in Britain. I. G. amarella, G. anglica and G. uliginosa. Watsonia 4: 169- 193. PRITCHARD, N. M. (1972). Where have all the Gentians gone? Transactions of the Botanical Society of Edinburgh 41: 279-291. PRITCHARD, N. M. & TutTIn, T. G. (1972). Gentianella Moench., in TutTin, T. G. et al., eds. Flora Europea 3: 63— 67. Cambridge University Press, Cambridge. Ricu, T. C. G. (1996). Is Gentianella uliginosa (Willd.) Boerner (Gentianaceae) present in England? Watsonia 21: 208-209. NOTES 113 Rose, F. (1998). Gentianella uliginosa (Willd.) Boerner found in Colonsay, v.c. 102. Watsonia: 22:114-116. Stace, C. A. (1991). New Flora of the British Isles. Cambridge University Press, Cambridge. R. L. GULLIVER Carraig Mhor, Imeravale, Port Ellen, Isle of Islay, PA42 7AL x AGROPOGON ROBINSONI (DRUCE) MELDERIS & D. C. McCLINT. The southern European and Mediterranean grass Polypogon viridis (Gouan) Breistr. (Poaceae) is now locally frequent as an established plant of damp rough ground in the Channe! Islands, and is sporadic in Britain on rubbish tips and similar places as a casual from wool, cotton, grain and probably other sources. There is a good drawing of it in Hubbard (1984, p. 304). It is distinct from other species of Polypogon in its entire (not 2-lobed), awnless (not awned) glumes, and was formerly included in Agrostis (as A. semiverticillata (Forssk.) C. Chr.). It is placed in Polypogon mainly on account of its spikelet disarticulation, which is well below the glumes rather than at the base of the floret. The species is readily recognized by its distinctive, rigid, dense, much-branched panicles, and by the minutely rough (i.e. strongly papillate) glumes. Sometimes it is accompanied on rubbish tips in Britain by another grass with a very similar habit and similarly papillate glumes, but with slightly notched glumes each with an awn up to 2 mm long. This is the hybrid Agrostis stolonifera L. X Polypogon monspeliensis (L.) Desf. (= X Agropogon littoralis (Sm.) C. E. Hubb.), which is also found as a rare native in southern Britain within the range of P. monspeliensis. The hybrid Agrostis stolonifera x P. viridis (= X Agropogon robinsonii (Druce) Melderis & D. C. McClint.) is a very rare grass, having been reported on only three previous occasions, all in Guernsey (McClintock 1975, 1987). It was discovered by F. Robinson in 1924 at St Sampson (N. E. Guernsey) and determined and named as Agrostis xX robinsonii (but written in error as X F. robinsonii) by Druce (1925), confirmed by J. Fraser and E. D. Marquand. The specimen is in BM, seen by C.A.S. The second record was made in 1953 at Vazon (W. Guernsey) (not at Grandes Rocques as stated by McClintock (1975)) by C. E. Hubbard (specimen in K, seen by C.A.S.), and the third in 1958 at Grandes Rocques (N.W. Guernsey) by D. McClintock, confirmed by A. Melderis (specimen in STP, seen by C.A.S.). In July 1994 P.M. discovered a single plant of a grass (herb. P.M.) that closely resembles P. viridis in habit in a long-abandoned industrial site in Shieldhall, Glasgow, Lanarks, v.c. 77. It grew 20 m from the edge of a lorry park in current use and c. 750 m from the nearest dock on the River Clyde. The habitat was scrubby grassland dominated by Holcus lanatus with a few plants of Agrostis stolonifera and very few of A. capillaris. No other grasses were present in the immediate vicinity. Other associates, one to a few plants in each case, were Artemisia vulgaris, Cirsium arvense, Chamerion angustifolium, Dactylorhiza fuchsii, D. X venusta, Epilobium montanum, Equisetum arvense, Luzula multiflora, Senecio jacobaea, Trifolium hybridum and Tussilago farfara. Scattered in the area were Salix caprea, S. cinerea subsp. oizifolia and S. X reichardtii, ranging in height approximately 1-3 m. The plant differs from P. viridis in its scarcely papillate glumes, lemma entire (not minutely toothed) at apex, palea c. 3/4 as long as lemma (not nearly as long) and anthers c. 1-2 mm (not c. 0-6 mm) long. The pollen grains are empty, and the plant is clearly a hybrid between P. viridis and a grass with non-papillate glumes, lemma entire at apex and much longer than palea, and anthers >1 mm long. Agrostis stolonifera fits this perfectly: glumes not papillate; lemma entire at apex; palea c. 2/3 as long as lemma; anthers 1-1-5 mm long. The Scottish plant closely resembles the 1924 and 1953 Guernsey specimens of X A. robinsonii. The glumes are not bifid and awned as stated by Sell & Murrell (1996) in any of these specimens. The 1958 Guernsey plant, on the other hand, is obviously a slightly unusual specimen of Agrostis stolonifera, being fertile and having all the diagnostic spikelet characters of that species. The Glasgow plant is therefore only the third known record of X A. robinsonii (Stace 1997). It must be considered a casual, having arrived at the site presumbly as hybrid seed, unlike the native Guernsey records. An intensive search of the site and surrounds was carried out by P.M. in 1996, but no further plants were found. Although the hybrid is endemic to the British Isles as far as is 114 NOTES TABLE 1. DIAGNOSTIC CHARACTERS OF X AGROPOGON ROBINSONII (TAKEN FROM THE 1953 GUERNSEY AND 1994 SCOTTISH MATERIAL) AND ITS PARENTS Agrostis stolonifera x Agropogon robinsonii Polypogon viridis Spikelet disarticulation below floret none near pedicel base Spikelet length (mm) 1-8-3 1-8-2-3 1-5-2-2 Glumes + smooth scarcely papillate conspicuously papillate Lemma length (mm) 1-3-1-8, c. 0-6-0-8 X as 1-2-1-5, c. 0-6-0-:7 X as__——0-7-1-0, c. 0-5—-0-6 X as long as glumes long as glumes long as glumes Lemma apex + entire, sometimes entire, awnless denticulate, awnless awned Palea length (mm) 0-8-1-2, c. 0-6-0-7 Xx 0-9-1-1, c. 0-7-0-8 x 0-6-0-9, c. 0-8-0-9 x as long as lemma as long as lemma as long as lemma Anther length (mm) 1-1-5 0-9-1-2 0-5-0-7 Pollen grains full empty full Caryopsis c. 1mm not formed c. 1mm known, it probably occurs in southern Europe where P. viridis is native, and whence the Glasgow plant might have been introduced. The diagnostic characters are listed in Table 1; see also Bradshaw (1975). REFERENCES BrapDsHAw, A. (1975). Agrostis L., in Stace, C. A., ed. Hybridization and the flora of the British Isles, pp. 579- 583. Academic Press, London. Druce, G. C. (1925). Agrostis verticillata Vill. X palustris Huds. (vel alba L.) nov. hybr. The Botanical Society and Exchange Club of the British Isles. Report for 1924 7: 457. HUvuBBARD, C. E. (1984). Grasses, 3rd ed., pp. 304-305. Penguin Books, Harmondsworth. McCuintock, D. (1975). The wild flowers of Guernsey, p. 267. Collins, London. McCuintock, D. (1987). Supplement to The wild flowers of Guernsey, p. 49. La Société Guernesiaise, Guernsey. SELL, P. & MurreELi, G. (1996). Flora of Great Britain and Ireland, p. 191. Cambridge University Press, Cambridge. Stace, C. A. (1997). New Flora of the British Isles, 2nd ed., p. 876. Cambridge University Press, Cambridge. P. MACPHERSON 15 Lubnaig Road, Glasgow, G43 2RY C. "Aa STACE Cringlee, Claybrooke Road, Ullesthorpe, Lutterworth, Leicestershire, LEI7 5AB GENTIANELLA ULIGINOSA (WILLD.) BOERNER (GENTIANACEAE) FOUND IN COLONSAY (V.C. 102), NEW TO SCOTLAND On 4 September 1978 the late E. C. Wallace and I were surveying various sites on Colonsay (v.c. 102) for a week. We found Gentianella uliginosa (Willd.) Boerner, new to Scotland, in the flushed ‘“‘machair’’ on the south facing slopes of Leac Bhuidhe, NW of Balnahard Bay, NM/426.004 on 4 September 1978. We were both quite convinced of its identity, but we sent specimens to the Gentianella specialist Dr Noel Pritchard who provisionally agreed’ with the identification but expressed a desire to see living material before fully confirming its identity. I was not able to revisit Colonsay until 1981, when I collected more material and sent it to him, and he has recently agreed that the Colonsay plants were good G. uliginosa. When I visited Colonsay in 1981 I also found G. uliginosa in machair type grassland at Traigh nam Barc (NR/355.909), and possible plants of it at Traigh Ban [Kiloran Bay] (NR/404.982) with much G. amarella. In 1978 we estimated that there were 30-40 plants at Leac Bhuidhe; in 1981 I counted 36 plants in the part of the site I studied. On 15 June 1989 I was surprised to see many plants coming into flower, but as most were not yet out, no proper count could be made, especially as rabbit grazing was then NOTES 115 TABLE 1. PLANT SPECIES OCCURRING IN A 1 M* QUADRAT WITH GENTIANELLA ULIGINOSA IN COLONSAY (V.C. 102) The first figure indicates cover and the second indicates sociability on the Braun-Blanquet scale (Shimwell 1971) Vascular plants Bryophytes Gentianella uliginosa 5 plants Ditrichum flexicaule agg. 2-2 Schoenus nigricans 3-2 Hypnum cupressiforme var. lacunosum 2-2 Festuca ovina agg. 2-2 Entodon concinnus 1-2 Lotus corniculatus Dae Pseudoscleropodium purum 1-2 Thymus polytrichus 2-2 Ctenidium molluscum +-2 Pilosella officinarum 1-2 Rhytidiadelphus triquetrus +-2 Bellis perennis 1-1 Trichostomum crispulum +—2 Euphrasia sp. 1-1 Linum catharticum 1-1 Plantago lanceolata j 1-1 Ammophila arenaria +-2 Campanula rotundifolia +-2 Plantago maritima +-2 Polygala vulgaris +-2 Centaurium erythraea oF Prunella vulgaris ae Species recorded outside the quadrat, but in the same community, included: Anagallis arvensis, Carex flacca, Galium verum, Gentianella campestris, Pinguicula vulgaris, Plantago coronopus, Selaginella selaginoides; with the bryophytes (in a wetter hollow): Cratoneuron commutatum subsp. falcatum and Drepanocladus revolvens; and the lichens: Diploschistes muscorum and Squamarina cartilaginea. Nomenclature follows Stace (1991) for vascular plants, Corley & Hill (1981) for bryophytes and Purvis, Coppins & James (1993) for lichens. severe, and many were bitten off. At Traigh nam Barc on 28 August 1981 B. J. Coppins, P. Wormell and I estimated that there were at least 40 plants. COMMUNITIES At Leac Bhuidhe on 26 August 1981, we noted that the G. uliginosa grew mostly on and around Schoenus nigricans tussocks in sloping flushes on blown sand on a south to south-east facing slope. A 1 m? quadrat was recorded on this date, Table 1. This vegetation probably equates to the Festuca rubra — Galium verum fixed dune grassland Prunella vulgaris sub community of the N.V.C. (Rodwell 1998, in press) but is more flushed and damper than the above community, with Schoenus nigricans, Selaginella selaginoides, etc., and approaches the character of dune slack vegetation in places. At Traigh nam Barc in 1981, the communities were rather similar but less flushed, and Schoenus nigricans was not seen at this exact location; however, Selaginella selaginoides, Antennaria dioica, and a rich bryophyte and lichen flora occurred on the damp calcareous machair, which at the time was a quite open community. It is worth noting here that, at the time of our visit, the machair areas of Colonsay were very fine, some of the best I have seen in western Scotland, and little disturbed by any human factors for a long time. ACKNOWLEDGMENTS Access to the sites was kindly granted by Lord and Lady Strathcona, Mr P. Hobhouse, Mr N. Grant and Mr A. McNeill. REFERENCES Corey, M. F. V. & Hit, M. O. (1981). Distribution of bryophytes in the British Isles. A census catalogue of their occurrence in vice counties. British Bryological Society, Cardiff. 116 NOTES Purvis, O. W., Coppins, B. J. & JAMEs, P. W. (1993). Checklist of lichens of Great Britain and Ireland. British Lichen Society bulletin 72: supplement. RopweELL, J. S., ed. (1998). British plant communities. Volume 5. Maritime and weed communities. Cambridge University Press, Cambridge (in press). SHIMWELL, D. S. (1971). Description and classification of vegetation. Sidgwick & Jackson, London. Stace, C. A. (1991). New Flora of the British Isles. Cambridge University Press, Cambridge. F. Rose Rotherhurst, 36 St Mary’s Road, Liss, Hampshire, GU33 7AH Watsonia 22: 117-125 (1998) Uy Book Reviews Welsh ferns, clubmosses, quillworts and horsetails, 7th ed. G. Hutchinson & B. A. Thomas. Pp. 265. National Museums & Galleries of Wales, Cardiff. 1996. Price £11.00. ISBN 07200—04—35-7. Welsh ferns is one of those books which has become a familiar old friend, parochially titled but broad in scope and which has developed and for the most part improved with the passing of each of its many editions. The latest, seventh, edition marks perhaps the most radical change, resulting in a book nearly half as long again as its predecessor. Gone are the hard cover and all mention of herbarium records. The major innovation, which I welcome, is the inclusion of maps giving Welsh distribution at the hectad (10-km square) level and thumbnail sketches of broader European ranges. The latter, while somewhat small, convey a useful impression. However the use of a mid-tone for uncertain occurrence is poorly reproduced and unclear. The introductory preamble on biology and morphology has been split, the morphological section moved until after the fern allies and now relating to the true ferns alone. I think this is a mistake and would have preferred an all-embracing section prior to the species accounts. The many nomencla- tural changes made in the 18 years since the last edition have been taken on board, although Asplenium trichomanes-ramosum L. should be called A. ramosum L., a name itself proposed for rejection. In some contentious areas, e.g. the taxonomy of Preridium, the authors give an overview without explicitly expressing an opinion. The treatment of all British taxa, which always made a nonsense of the volume’s title, has been taken to an extreme, with all taxa, however briefly naturalised, getting a mention and a token Welsh name. The manufacture of vernacular names attracts strongly polarised views; I dislike it and it adds here to an already cumbersome and less than easy to use index! The original half-tone plates have been re-photographed and are deplorable. There are niggling errors aplenty. Cystopteris alpina, an extinct(?) native, is curiously completely omitted. Asplenium viride is given as L., not Huds., throughout. A. x badense was convincingly shown to be an aberrant, possibly octoploid, A. ceterach and not of hybrid origin in 1989. Herbarium specimens of reputedly British A. fontanum do exist, etc. The world distributions for some taxa have been retained from earlier editions although taxonomic changes have occurred in the meantime, e.g. we have the near cosmopolitan distribution of Trichomanes radicans given instead of the true Macaronesian-European endemic range for T. speciosum. Similarly Huperzia selago does not occur as stated in Macaronesia where it is replaced by H. suberecta. In spite of these criticisms there is much to recommend this inexpensive and useful guide, which like its antecedents deserves a place on the bookshelf. F. J. RUMSEY The botanists and guides of Snowdonia. D. Jones. Pp. 174. Gwasg Carreg- Gwalch, Llanrwst. 1996. Price £6.95. ISBN 0-86381-383-6. Obtainable for £7.50 (incl. p. & p.) from Gwasg Carreg Gwalch, 12 Yard yr Orsaf, Llanrwst, LL26 OEH. This smal! softback book is described as “‘an account of the botanical exploration of Snowdonia from the earlier period when the Herbalists and Apothecaries conducted ‘simpling voyages’ into the countryside to gather plants for medicinal uses, up until the Victorian era by which time botany had developed into a separate science’’. Having written comparable accounts in Floras of Shropshire and Montgomeryshire (including tracing different activities of many of the same botanists), I well understand the difficulties of providing readers with a satisfying balance between biographical information, anecdotal material and sometimes long lists of the species found and of maintaining some thematic unity in the work as a whole. 118 BOOK REVIEWS Dewi Jones breaks up his account into 19 chapters, but the guides who led visiting botanists up Snowdon and some of the local botanists, as well as the routes they followed, often serve to link these together. On the historical side (as Gwynn Ellis mentions in his foreword), particularly enlightening is the underlying theme of the contrast between the life style of the English gentry who wished to see the sights and plants of Snowdonia and that of the local peasantry who guided them. I found that the botanical information did not always come to life quite as vividly, but a full index to plant names enables one to trace the story of individual species, sometimes over nearly three centuries. I tried this with Lloydia serotina (18 entries, from Edward Lhwyd’s discovery of it, published in 1696, to the splendid story of the 19th century Oxford Professor in smooth-soled leather boots who never saw the Snowdon lily in situ on the Glyder cliffs because he was too terrified to open his eyes), Saxifraga nivalis (eleven entries) and the two Woodsia species (nine entries jointly) and gained a perspective on them which I had not obtained when first reading the book. However, the result was far less satisfying with Saxifraga cespitosa (four entries). I can scarcely fault the botanical index, but too many people mentioned in the text do not appear in the general index. Twelve Robertses and ten Joneses are indexed, but I have found two more of each in the text! The choice of works in the bibliography seems arbitrary, and some of the authors of works mentioned in the text but not in the bibliography are in the index, some not. The twelve colour photographs of plants are satisfactory, but some of the 28 black-and-white illustrations much less so. I wish I knew the sources of the coloured print of the busy summit of Snowdon in the 1850s on the cover and of the monochrome print on the title page. Jones does not stick strictly to Snowdonia in this book. The first chapter, on the 16th century manuscript herbal of William Salesbury (whom I prefer to spell Salusbury), is included on the grounds that its author “spent the greater part of his life at Plas Isa, Llanrwst’, and another chapter is about Hugh Davies’s Welsh botanology (1813), essentially a Flora of Anglesey with an alphabetical list of Welsh plant names with their Latin and English equivalents. My editorial fingers itched at the erratic punctuation and misspellings; some of the Latin is faulty too. But in the end I was prepared to follow a few byways and to tolerate some bumps along the road. This is a book packed with fascinating information about North Wales, historical, social and botanical. P. H. OSWALD The introduction of Chinese plants into Europe. L. A. Lauener, edited by D. A. Ferguson. Pp. xii + 269. S.P.B. Academic Publishing, Amsterdam. 1996. Price Dfi. 140.00/US $87.50. ISBN 90-5103— 130-0. When Andrew Lauener died in 1991, he left a virtually complete manuscript of this book, the fruits of many years’ experience of working with Chinese collections in the Herbarium of the Royal Botanic Garden, Edinburgh. David Ferguson, with the help of some colleagues, prepared it for publication, which was facilitated by friends in The Netherlands. The author approaches his subject mainly from a systematic angle, unlike Emil Bretschneider’s monumental History of European botanical discoveries in China (1898), E. H. M. Cox’s Plant hunting in China (1945) and Roy Lancaster’s Travels in China: a plantsman’s paradise (1989), the longest chapter by far being that entitled “‘The plants’’. This consists of an alphabetical account by genus of how some Chinese plants reached our gardens. The factual details (often fascinating) of collecting and introducing them into cultivation are accompanied by comments about botanists who have worked on the genus and their publications, about the names, nomenclatural history and botanical features of certain species, and about their uses both in their native region and in cultivation. All this information is provided in a personal way, reflecting the author’s experience and including remarks on who is working on what and what is due to be published soon (data that are inevitably sometimes out of date). Each genus is illustrated by one or more bold line drawings that give a good impression of the plant in question but may on occasion strike some readers as being rather dark, even fuzzy. The book also includes small chapters on China, Chinese place names and plant collectors (alphabetically), a selective bibliography, a useful gazeteer of most place names mentioned (with co-ordinates) and indexes. Readers interested in the activities of botanical collectors in China will find little in this book that BOOK REVIEWS 19 has not been treated in more detail in the three works mentioned above (only the first two are in the bibliography). What it will do is answer such questions as “Who introduced that species into cultivation?’ and, for some species, ‘Has it always been grown under that name?” A certain amount of apparently irrelevant information, a comment about Rubus chamaemorus, R. idaeus and R. fruticosus, for example, would seem to be aimed at gardeners who are more familiar with European members of the genus than with Chinese ones. In general, however, the information given is usually accurate, though sometimes incomplete. If you know that a garden plant is Chinese in origin, you are likely to find some interesting and/or useful background data about it in this book. N. K. B. ROBSON Aquatic plants in Britain and Ireland. C. D. Preston & J. M. Croft. Pp. 365. Harley Books, Colchester. 1997. Price £25.00. ISBN 0—946589-55-0. The introduction clearly states that this book summarises the distribution, habitat and reproductive biology of the vascular plants which grow in freshwater in Britain and Ireland. For the purposes of identification one is referred to Stace’s New Flora of the British Isles (1991) or to special literature. I find it a pity that there are no keys or “important diagnostic characters’; there is rather a lot of unprinted paper in the book and it would not have been longer or more expensive with, at least, some information on critical taxonomic features. In some cases this is important because new introductions such as Cabomba caroliniana and Hydrocotyle ranunculoides are not described in Stace’s Flora. Granted, both species are illustrated but the illustrations are without scales and are not much more than habit sketches; they do not allow critical determination. It is no doubt unfair to criticise something which is clearly stated not to be an aim of the book but its title and very attractive cover may lead some people to believe it is a “popular” work. This is a serious scientific work written for experienced botanists. The authors discuss the problems concerning the definition of what an aquatic plant is. They attempt to include those species which “‘characteristically”’ grow in water which persists throughout the year. In the Apiaceae some species which are deliberately excluded are listed including Sium latifolium; this species germinates under water and develops finely divided submerged leaves and it Overwinters under water. This book is for summer botanists. Apium repens and its hybrid with A. nodiflorum are not even mentioned; in my experience the hybrid is rather more aquatic than, for example, Calla palustris and Myosotis scorpioides which are included. Hypericum elodes, at least in Ireland, may grow in water the whole year but it is not even mentioned, neither are Cotula coronopifolia, Lycopus europaeus, Teucrium scordium and Samolus valerandi. The choice of species is a problem of personal experience and is not important — all the “real” aquatics are included — but it would be useful to know which “‘aquatic” species were intentionally excluded. Each genus but not each species is illustrated. The purpose of the illustrations in an important scientific book like this is not quite clear; some of them, such as Cabomba caroliniana and Lagarosiphon major, are poor. Each species has an updated distribution map clearly showing Hee in the distribution. Distinct symbols are used for pre-1950, 1950 to 1969, and 1970 and later records. It is very depressing to see the changes since 1950. The introduced species have different symbols and, on the whole, they seem to be doing rather better than the natives. An interesting case is Sagittaria latifolia which seems to be slowly taking hold in the south. Sagittaria sagittifolia is a very distinct plant: everyone knows what it looks like and it is rarely gathered. Nobody needs to poke into its flowers or put its fruits under a lens. However, if you do it may turn out to be S. /atifolia. Around Ziirich it seems to have virtually replaced S. sagittifolia and nobody saw it happen! For this reason it would have been valuable to have some diagnostic characters; the authors do, at least, say it may be overlooked because of its similarity to S. sagittifolia. The habitat descriptions are very concise and clearly presented, also they are well documented. The bibliography runs to 24 pages! The part on reproductive biology is rather mixed: for some species it is excellent but for others almost no information is given. It would have been valuable to know more about which species are self-incompatible and the vectors for pollen transfer among insect-pollinated plants. The nature of the effective disseminules and their dispersal could have 120 BOOK REVIEWS been expanded; the authors have considerable experience with these plants and should, perhaps, have more often added their own observations rather than relying on published work. Conservation designations of the rare species are presented at the end of the book. I have devoted most of my life to the biology of aquatic plants. This book has delighted me and I have learnt a lot of new information. What more can one wish? This is an essential work for all people concerned with the ecology and management of freshwater. The standard of production is very high — I failed to find any serious mistakes. For a book of this quality and length the price is remarkably low. C. D. K. Coox The making of the Cretan landscape. O. Rackham & J. Moody. Pp. 237. Manchester University Press, Manchester. 1997. Hardback £50.00, ISBN 0—-7190-3646-1. Paperback £19.99, ISBN 0— 7190-3647-X. Crete is the most southerly region of Greece and the largest, most mountainous Greek island. The native flora is of enormous national and global importance, and one of the choicest in the plant-rich Mediterranean region. Of some 1650 higher plant species, 10% are endemic or shared only with Karpathos and Kasos to the east. This unique and ancient flora, as significant as that of most tropical islands, is also irresistable to the hundreds of British botanists and naturalists who flock there each spring. Now a book is available to supplement their field guides. In The making of the Cretan landscape, English botanist and landscape historian Oliver Rackham and American archaeologist Jennifer Moody present the fascinating and complex story of the vegetation and landscape of this beautiful island. They dedicate just one chapter directly to the flora and two to the vegetation, but their holistic approach is essential to a full understanding of the processes that have influenced, and continue to influence, the varied plant life of Crete. The book is beautifully written, scholarly but never losing its humour or lightness of touch. This exciting work, which pursues themes familiar from Oliver Rackham’s 1986 classic, The history of the countryside, will guide the curious visitor, whether botanist, conservationist or bemused tourist, through the complexities of Cretan vegetation and landscape. The authors break away from the conventional philosophy and restrictive practices of contemporary biology and ecology to combine their field observations, trawls through Venetian archives and broad perspective of geography and history. They take the reader from the earliest geological and prehistoric beginnings, through the ““Golden Ages” under the Minoans and the Byzantine Empire, superseded by long Venetian and Turkish occupations, to the modern Crete of E.U. subsidies. On Crete, links with the past are real and tangible. Agriculture has long been the basic occupation of the people, and the prosperous market town of Archanes near Iraklion has remains of a Minoan farm, complete with olive press. Rackham & Moody cite the observation of a visitor from Renaissance Venice who recorded “‘Iusgriano con fior d’oro”’ (Golden Henbane, Hyoscyamus aureus) on a bastion of Iraklion’s huge 16th century city walls. This predominantly south-west Asian plant survives today at the site. The opening chapters, on geology, physical geography, climate and animal life past and present, include a useful account of how the famous flower-rich gorges and the distinctive flat mountain- plains like Lassithi and Omalos may have formed. The simple, elegant map of gorges and mountain- plains in the preface is the only one of its kind in published form. Oliver Rackham is a fine cartographer and calligrapher and his diagrams and maps embellish the text throughout. He provides valuable distribution maps of major endemics, such as Cretan Wall-lettuce (Petramarula pinnata), Cretan Sainfoin (Ebenus cretica) and Cretan Dittany (Origanum dictamnus), and the principal trees, among them the endemic Ambelitsia (Zelkova cretica) and Cretan Date-palm (Phoenix theophrasti), the latter now known also from the Dodekanisos and adjacent Turkish coast. The maps illustrate new data, notably Hungarian Oak (Quercus frainetto), otherwise not recorded from Crete. Another oak, Q. brachyphylla, is usually recorded as Q. pubescens. One should note that Rackham & Moody eschew (probably wisely) recent taxonomic progress, including the revised Flora Europaea Volume 1! Above all, Rackham & Moody endeavour to dismantle the simplistic but widespread notion that Crete, and much of the Mediterranean region, is merely a ruined landscape or “Lost Eden”. The BOOK REVIEWS 121 conventional view is that grazing by goats and other domestic stock is the culprit. The authors refute this, stating (their italics) that ‘“There can be no doubt that in Crete, ‘excessive’ browsing is not an artefact, but is the natural state to which the flora, and especially the endemics, are adapted.”’ While native vegetation has certainly been profoundly modified by more than five millenia of human activity, the authors argue that today’s vegetation, notably the scrublands and woodlands, may actually be in better condition than for centuries. This has immense implications for nature conservation in Crete, the rest of Greece and perhaps all the lands around the Mediterranean. The authors pursue their radical hypothesis with reference to flora, vegetation and landscape history in the context of the socio-economic and historical background of human settlement, trade, rural industry, roads and tracks, animal husbandry and vernacular architecture. The Cretan flora itself remains a dynamic, living resource, with wild plants still being utilised by the local population. The authors argue that, despite often considerable fluctuations in tree cover and agricultural priorities, basic ecological patterns and processes remain the same. Indeed the ancient Minoans and Mycenaeans may well have known similar vegetation to what we see today. Overgrazing and bulldozing of new olive terraces is certainly a problem in many areas, but once the visitor starts to look around, it is remarkable how much land is covered by impenetrable scrub and stands of woodland, wood-pasture and what the authors rightly call savanna. Rackham & Moody are undoubtedly closer to the mark than the Greek pundits who see Crete’s apparent lack of forest cover as another calamity of Turkish rule! Rackham & Moody demonstrate convincingly that most of the woody vegetation has either persisted or regenerated over the last 150 years. Landscape drawings of 19th century visitors such as Edward Lear in 1864 show this clearly. Pines, cypresses, maples and oaks have invaded former fields and terraces, both in the lowlands and the mountains. In western Crete, high maquis dominated by Strawberry Tree (Arbutus unedo) and Tree Heather (Erica arborea) has developed into dense woodland. This echoes a trend seen over the rest of Europe in the late 20th century, as people drift away from the countryside and land falls out of cultivation. In the final chapter, ‘““Conservation and the future’’, the authors are more optimistic than most commentators. Nevertheless, they are concerned about current trends in the Cretan landscape. They note especially ribbon development on the coast, destructive bulldozing for new roads and olive terraces, and “the idea that the irrigation of olives is a necessity rather than a bonus. The search for water has dried up rivers and springs. . . All this is for a subsidized monoculture unlikely to last long.” Certainly, arable weed communities such as those in small upland fields dominated by bulbous or tuberous-rooted, arable weeds like wild tulips and Geranium tuberosum, are threatened by progress. As all over Europe, wetlands have disappeared or diminished, although coastal lagoons and marshes survive here and there. Fortunately, apart from plants of coastal sands such as Androcymbium rechingeri and Centaurea pumilio, relatively few endemic plants are threatened by habitat destruction. Many are under threat due to their natural rarity, but they grow in remote places with few or no visitors. Most occur on remote rocks and screes, high up on the sides of gorges or precipitous cliffs; several, like Cretan Wall-lettuce and Verbascum arcturos, thrive on walls. The authors rightly conclude that the way forward must lie in education, and are cheered by the fact that younger people in Greece are waking up to the interest, value and fragility of their native environment. They stress the contribution of ecotourism, since increasingly people visit Crete for its natural history. It is certainly a way to restore prosperity and people to village communities in the mountains. The island has breathtaking, dramatic scenery, fine antiquities and excellent tourism facilities, and not least traditional Greek xenophilia or hospitality to travellers (these days, alas, tempered by a canny, sometimes aggressive commercial streak). Even ecotourism can damage fragile plant, animal and human communities, but it is much less a threat than outright habitat destruction through uncontrolled development. Oliver Rackham and Jennifer Moody provide much food for thought, and a good read on the long flight to Chania or Iraklion. Their splendid book will provide the thinking naturalist - why do so many amateur botanists merely want names for plants? — with the indispensible background — information to appreciate the Cretan flora. One hopes too that at least the odd copy of The making of the Cretan landscape finds its way on to the desks of decision makers and politicians in Greece. The book is a triumph, a milestone in the study of the Mediterranean world and its ecology. J. R. AKEROYD 22 BOOK REVIEWS Flora of Great Britain and Ireland. Volume 5. Butomaceae—Orchidaceae. P. D. Sell & G. Murrell. Pp. xxi + 410. Cambridge University Press, Cambridge. 1996. Price £60.00. ISBN 0-521—55339-3. In the light of destruction of forests for paper, one has to ask with the publication of the third major Flora of the British Isles by Cambridge University Press in ten years, is it worth having? My feeling is an emphatic yes! Chop down those forests — this is a landmark in British botany. The Clapham, Tutin & Warburg era can now be laid to rest with the highest honours and affection. The Flora of Great Britain and Ireland has largely been written by Peter Sell with the assistance of Gina Murrell and Potamogetonaceae and Ruppiaceae contributed by C. D. Preston. The aim of the Flora is to supply full descriptions of all the species in Stace’s (1991) New Flora of the British Isles, to include all the large apomictic genera and many infraspecific variants, and to add more information about hybrids. It is thus a detailed desk-top reference work complementary to Stace, the latter concise and portable. The Flora follows the classification used in Stace’s New Flora (1991) and Kent’s List of vascular plants of the British Isles (1992), resulting in a degree of harmony between the three, with the genera and species sometimes in a different order (Pontederiaceae is listed under Bromeliales in the Conspectus but in the correct place in the text). There are some taxonomic differences from Stace and Kent such as the plant hitherto called Zostera angustifolia included under Z. marina, Carex viridula under C. flava, and Dactylorhiza lapponica under D. traunsteineri. As a result of the extensive nomenclatural work done on the British flora in recent years there are few changes in names, and there is no difficulty in relating the three due to the extensive synonomy. As the Flora aims to provide full descriptions I checked those for the first species on or after pages 1, 90, 180, 270 and 360 against our herbarium material in the National Museum and Gallery of Wales (NMW). In general the accounts all worked well though there were the few usual specimens and measurements exceeding the stated limits, perhaps because we grow larger and better plants in Wales than in Cambridge. The bracteole-like structures present in the inflorescence of Butomus umbellatus were not mentioned. Our plants of Carex diandra are significantly larger with stems to 75 cm and leaves to 60 cm X 4mm. The account of Deschampsia flexuosa is very good. We have no material of Chionodoxa luciliae and I wonder if it really merited inclusion as it has only been recorded from lawns of Cambridge Botanic Garden. I was surprised to find Ophrys insectifera has leaves up to 8 cm wide — ours are rarely more than 2 cm. Overall there was a refreshing feeling that the measurements and descriptions had been looked at afresh and not copied from elsewhere. The keys to species include infraspecific taxa as well as the species, and those to Juncus, Luzula and Puccinellia | checked worked well. A few literature references such as the Biological Floras or taxonomic papers are cited. Summary distribution data for all taxa are provided. Inevitably a few are out of date (e.g. Cyperus fuscus has been known in Berkshire since the early 1980s) and later volumes will no doubt draw on Atlas 2000. Judgement on treatment of critical groups will really have to wait for the Hieracium treatment in volume 4, and the only really critical genus dealt with is Dactylorhiza. The genus has a novel treatment, probably the most complex in the Flora, with numerous subspecies, varieties and forms, and useful details of the hybrids. I found it difficult to apply to herbarium material due to the necessary emphasis on flower features, but the detailed accounts will help interpret variation present in the field. The drawings of the labella will help, and I would have liked colour photographs. A substantial amount of the book is dedicated to the intraspecific taxa, and this Flora provides the first recent serious attempt to summarise the information available. This was my only disappoint- ment, not because of what was included but because I was left wanting to know more. There is no means of tracing any infraspecific taxon not included in the Flora, even if they are no longer worthy of recognition. For instance, we have many specimens of Deschampsia flexuosa in NMW named as var. montana (L.) Hook. f., a variety mentioned in Hubbard’s Grasses (1984), and I was left wondering if it existed or not, and similarly what has become of Carex diandra var. major Koch? Perhaps these are our big Welsh plants?. Many infraspecific taxonomic problems remain to be investigated. Are there Norfolk plants of Alisma gramineum subsp. wahlenbergii or not? There are specimens of Juncus maritimus var. atlanticus from Cornwall and Hampshire in BM (and possibly elsewhere), and a similar plant occurs down the west coast of Europe to at least northern Spain where it appears to have been described as BOOK REVIEWS 123 var. longipedicellatus Sen. & Elias. Similarly, a form of Juncus maritimus visually identical with var. congestus occurs in Spain where it has been named as var. compactus Elias. Jan Kirschner and I do not accept Luzula multiflora subsp. frigida as a British plant, though the material of Luzula multiflora from Scotland would repay further study (our 1996 subsp. hibernica was published too late for inclusion). I am uneasy about the way infraspecific taxa in many genera have been strait- jacketed into one infraspecific rank, but thankful that there are few new combinations as a result. These problems simply point to the huge amount of work which remains to be done on infraspecific taxa, and this Flora gives an excellent baseline from which to direct further work. The full descriptions of aliens will be helpful to botanists without ready access to literature or herbaria. An alien in the hand can now be checked in detail against the description — all too often the brief accounts in some Floras result in an assumption that the plant in the hand is the one in the book. Deciding which aliens to include and which to exclude presents its usual difficulties; the approach taken is to include as many as possible. Some species have been given fuller treatment than others (compare on the same page the full account of Chionodoxa luciliae, with the briefly mentioned Muscari azureum, apparently naturalised in derelict parkland in Somerset). On the basis of the information available to date, both Serapias lingua and S. parviflora merited full treatment as native species. The book has a laminated hardback cover with a picture of bluebells. It contains an introduction, a conspectus of families for the whole work and a key to the monocotyledonous families included, the systematic accounts, a separate list of new taxa and combinations, a glossary and an index. The typeface and layout is that of the third edition of Clapham et al., Flora of the British Isles (1987) and is not distinctive, and the illustrations are somewhat functional. There are a few minor formatting and typographical errors. To sum up, the two key features of the Flora are the full descriptions of British plants with up- dated nomenclature, and the treatment of the infraspecific taxa. It is an essential reference work for academic and serious amateur botanists in the British Isles, and probably for all in north-west Europe. This is a taste of Peter Sell’s outstanding, detailed knowledge of the British flora after a life- time of study, and I sincerely wish Peter and Gina the best of luck with its completion. T. C. G. RIcH New Flora of the British Isles, 2nd ed. C. A. Stace. Pp. xxx + 1130. Cambridge University Press, Cambridge. 1997. Price £28.95. ISBN 0-521-58935-5. It is five years since Arthur Chater reviewed the first edition of this book in Watsonia 19: 161-163. As predicted it has become the standard Flora for taxonomy, nomenclature and identification within the British Isles. The limited and idiosyncratic typeface and skeletal index of the first edition have now been thoroughly remedied. The index now occupies 131 pages and the new typeface is clear, more varied and less wasteful of space. As a consequence, on roughly the same number of very slightly enlarged pages, a more generous margin is provided at the foot of each page and an additional 320 taxa are included. The book is bound in similar flexible plastic covers to the first edition. Despite its weight, just exceeding 1-5 kg, the bindings appear to be adequately robust. I have received but one report of the binding of the first edition failing. Although the plastic cover is clearly designed to improve its durability in the field, the weight and bulk of the book are a severe drawback to it becoming a regular companion. As an aid to its use in the field, a centimetre and millimetre scale is provided on the inside front cover. In this edition its length has been usefully extended but it has been placed so close to the edge that the millimetre scale is barely visible on my copy. Taxonomic and nomenclatural changes between editions have been mercifully few. I am particularly pleased that the name Asplenium viride has been restored to us, together with Drosera anglica and Fragaria moschata. On the other hand Helianthemum canum becomes H. oelandicum, — to join a select band of half a dozen taxa or so now sporting new names. Huperzia selago has been split into two subspecies, whilst Arctium has been completely revised, following H. Duistermaat, to now recognise three species only. This latter genus is still described as “difficult”, which on first use of the key and new descriptions cannot be disputed. Other minor changes have been made to the 124 BOOK REVIEWS treatment of subspecies, e.g. within Luzula multiflora. Elsewhere many of the larger keys have been subject to minor changes to improve their performance and take in additional taxa. Cotoneaster, for example, has been extended from 45 to 68 species. The treatment of hybrids is still very uneven. Few generic keys (the most notable exception being Potamogeton) include hybrids. Some are described in detail whilst only the British distribution of others is mentioned. The generally sparing treatment of intraspecific taxa may also still not suit all tastes. No attempt has been made to extend the ecological notes which are still far more brief than those of Clapham, Tutin & Moore (1987) in their Flora of the British Isles. Nor is any information provided on world distribution except for introduced taxa. Chromosome numbers have, however, been added to the second edition and distribution information mostly brought up to date. Technical terms have, according to the preface, been further reduced. This is to be welcomed and should be carried further in future editions, perhaps also considering an expansion of the glossary. In this, as in the first edition, no doubt to save space, users of the glossary are referred back to the generic texts for an explanation of some terms. This is not helpful when the latter description has forced reference to the glossary in the first place. Improvements have been made to the illustrations. The small unsatisfactory silhouettes of Euphrasia have been replaced with excellent line drawings. Examples of species from the major sections of Hieracium, Taraxacum and Rubus are now included as line drawings for the first time, as are illustrations of Polypodium sporangia. Unfortunately no changes have been made to the illustrations of Odontites, Veronica hederifolia and Ranunculus omiophyllus criticised by Chater in his review of the first edition. In compensation other numerous small improvements include the addition of numbers to the family and generic names on the page headers, making recourse to the now formidable index largely unnecessary when moving from one key to another. Taken together, the typeface change and large number of other small changes and improvements render this new edition so much more pleasant to use that, even if you already have a first edition, I strongly recommend that it is worth purchasing the second edition. It is still excellent value for money and whilst the native flora may not be covered in such depth as in other Floras, this is more than made up for by the large number of non-native taxa not covered elsewhere so conveniently. As undoubtedly the standard Flora for the British Isles for years to come, it is to be hoped the publishers will consider the investment worthwhile to ensure that the quality of the illustrations in future editions matches the quality of the text. R. G. Woops A dictionary of plant-lore. R. Vickery. Pp. 437. Oxford see ey Press, Oxford. 1995. Price £14.99. ISBN 0-19-866183-5. Anyone accustomed to the orderly, cut-and-dried world of field botany is likely to experience a series of shocks on any extended acquaintance with the literature of the study of folklore. For a start, there is no consensus on where that subject begins and ends. Many folklore collectors, for example, ignore medicinal uses of herbs, apparently regarding these as too mundanely practical to count for their purposes and perhaps better left to their kinsmen, the students of folk life. Then there is the lack of large-scale compendia, on the lines of Topographical botany or the Biological Flora of the British Isles, which would enable one to tell what is known already. In the absence of such works much energy is wasted in recording the same things over and over again. Only rarely, moreover, is information quantitative or spatial, making it impossible to tell how many people in any one locality or area follow the belief or practice in question or how widely it is to be met with in terms of geography. Newcomers will further find to their horror that few books on folklore are indexed, so there is usually no alternative but to search through them page by page. For having the persistence to operate on such a dauntingly unsatisfactory front the author of a volume like the one under review deserves both sympathy and admiration. To a large extent, though, Vickery has sidestepped the worst of the problems by restricting his concerns in this particular case mainly to items recorded in the period 1981-94, just to the British Isles and just to plants (wild and cultivated alike). Essentially the book is the fruit of information provided, unsystematically, by some 700 informants. There is no means of knowing how far the reports BOOK REVIEWS 125 received were representative of the present-day pattern of behaviour overall, and all that could be done was to select some of the reports for quoting, simply as illustrations of the variety of beliefs and practices attached today to a great range of plants even in such a comparatively sophisticated set of cultures as those to be found in these islands. Many of the quotations are quite lengthy and make colourful reading. Within these limitations the coverage is comprehensive, and the average reader should find his or her curiosity satisfied on just about every matter likely to raise a query. The fact that the author is an experienced botanist in addition to being one of Britain’s leading folklorists gives this work a special authority, and the punctilious way in which the source is cited for every particularised piece of information is assurance enough of the scholarly standard that obtains throughout. Handy in size, attractively produced and very reasonably priced, this is a work which many readers of this journal will want to have not only on their shelves for reference, but also on their bedside tables for occasional dipping into — worthy successor as it is to those long-prized Victorian standbys, Mrs Lankester and the evergreen Anne Pratt. D. E. ALLEN Dh AN, i cosaclon otf omnes * elaine tae t S Ss x Wa: event You @ Ae 1s ae ae) ti Ah Lets (i Aasdy Wena shen Oita ow re suet dnt ) rH Gib hy > CNS TA COV Cnt i pow We TOR VERES Ot t ; J mt ha * vent Gti bees Man Ii wi agate that ae : a! aoteee a " 0 i i j iy , " i, a i ie: Ry Oe eee eae NABER geta enie ey aa Tama pil it) Py ahs ey Re ve Fehon, i PP Obs \ ‘ he ae ! ’ mt eee at, FLV oe : } aff me Ry SR RRS ue). eRe SRM Le AHA ON Te Me ime Oe a al Bsa a Ae 4 J Nt aA j Ba a «A ™. ie ‘a Pity (A es ORR ae ar ah Ww Roa Me oe) yy he pa (hy cy ic ya m mh bd, nan wt ai i ity ee i F tu b 4 i ine i { 1 RIA / > STAC VUE RUM bh Rc Ar AL Has i f fui Ain ay ae ee | i el A aa i Site cenit, enn Hie ‘ , ¥ iy as it pant , r ha ‘ ‘3 enh UR ei ete, hb rer Watsonia 22: 127-138 (1998) 129 Obituaries DOREEN SYLVIA LAMBERT M.B.E. (1915—1996) Doreen Lambert died unexpectedly in July 1996. She had been for many years the B.S.B.I. Recorder for both Co. Londonderry (v.c. H40) and Co. Tyrone (v.c. H36), a job which she pursued with characteristic vigour and attention to detail. Originating from Co. Galway, from an old Anglo- Irish family of the sort which at times seems more English than the English, despite generations spent in Ireland, Doreen claimed her family was connected with Colonel Lambert, one of Oliver Cromwell’s notable lieutenants who settled in Ireland during the Commonwealth. The Lambert clan became well established in the Oughterard district of Galway, but Doreen and her parents moved to Garvagh in Co. Londonderry in the 1930s, where Doreen used to “‘shoot the hills” with her father. During World War II she characteristically volunteered for service in the British Forces, becoming an officer in the Royal Air Force, attached to the newly developing Radar arm. During part of the war she was stationed at a Radar station atop the Great Orme at Llandudno, an excellent posting for a field botanist as she fondly remembered! She was awarded the M.B.E. for her wartime services. Later she pursued a successful career as a civil servant in Northern Ireland, latterly working in the SNC Monuments section of the Department of the Environment, Northern Ireland. After retirement in the 1970s she and her mother (an astonishing lady in her own right who died at the age of 102 with a mind as clear as a bell to the end; like many of the Irish she had a great fondness for horses) moved to the small seaside resort of Castlerock on the windy north coast of Co. Londonderry (Doreen always, incidentally, referred to it as ‘“Derry’’), to a large bungalow with magnificent views of the Atlantic Ocean from the picture windows. There she settled down in earnest to pursue her hobby of field botany which had begun before the War; she had gone out botanising in the 1930s with Robert Lloyd Praeger, and had become very friendly with the late Miss Pat (“‘Paddy’’) Kertland of Queen’s University, Belfast, who was the leading light in field botany in Northern Ireland in the 1950s and 1960s, and with Maura Scannell, of the Botany Dept at the National Museum in Dublin, later of Glasnevin. I first met her in the very early 1970s at Miss Kertland’s home, and from about 1973 we developed a mutual passion for dandelions inspired by John Richards’ Taraxacum Flora published in 1972. Many of the first Irish county records of various Taraxacum species are Doreen’s. She collected assiduously, not only in the northern counties, but also in her native west of Ireland, where she was able to show that Taraxacum palustre and its allies were nowhere near so rare as had been hitherto thought. 1972 was also the year that Miss Kertland produced a Supplement to Praeger’s 1938 edition of the Flora of the north-east of Ireland, helped considerably, as the title page states, by Doreen. From about 1976, by which time Doreen had settled down at Castlerock and had become the Recorder for Cos Londonderry (H40) and Tyrone (H36), we worked together on producing a revision of the Flora of the north-east of Ireland. Doreen, with her collaborators in Co. Londonderry, notably Ian and David McNeill, John Harron, David Riley and Mrs Jo Newbould, provided me with a steady, and at times almost overwhelming, stream of records, sightings and queries. At the same time she was vigorously opposing the spoilation of the countryside of her adopted county, especially of its beautiful coast. A keen golfer, she used to botanise on the links at the same time as golfing, but she vociferously opposed the damage done to the north coast dunes by extensions of local golf courses at Castlerock and Portstewart and was instrumental in averting the worst consequences that might have followed without intervention. Her golfing activities included membership of the Castlerock Golf Club, on which course she had seen what she considered to be Campanula giesekiana. Determined to show this to me one day, she took me on to the course and walked past and through small armies of irate golfers who kept shouting warnings at us. I expected any moment to be struck down by a golf ball, but Doreen placidly ignored the shouts, said “I’m a member!” and leisurely proceeded on her way. 128 OBITUARIES She was a careful and meticulous observer and was quite an accomplished botanical artist — she recorded many of her more notable finds as watercolours in a series of loose-leaf albums. It includes carefully observed drawings of most of the Taraxacum microspecies that she had found and collected; together with her herbarium material the collection of drawings is housed in the Ulster Museum, Belfast. Doreen’s contribution to our botanical knowledge of the north of Ireland is considerable, as any glance through our 3rd edition of the Flora of the north-east of Ireland, which appeared in 1992, will show; her name appears with the other principal collaborators on the title page. (Whilst engaged on the Flora of the north-east, she was simultaneously preparing a typescript draft Flora of Tyrone.) It was my privilege to know Doreen as a friend for nearly 25 years, and I was saddened by her unexpected death, but whenever I visit the north coast dunes I will half close my eyes and remember her stocky, jovial figure, the upper class accent, and long talks about hybrids and subspecies over gin-and-tonic with the majestic waters of the Atlantic in the background. P. HACKNEY EDGAR W. B. H. MILNE-REDHEAD M.B.E, L.S.O., T.D. (1906—1996) Edgar Wolston Bertram Handsley Milne-Redhead, who died on 29 June 1996 in his 91st year, was a professional systematic botanist but better simply described as an all-round naturalist; field work had more appeal for him than writing monographs in the herbarium. He will be best remembered as an ardent and successful conservationist, a collector of superb specimens, for his editorial work but perhaps above all for firing the enthusiasm of many amateur botanists, particularly in Africa, to | | | } | } Se Se OBITUARIES 129 emulate his own magnificent results. He had at first sight an austere military air caught to perfection by his Belgian colleague Prof. J. Léonard (a co-founder of Association pour l’Etude Taxonomique de la Flore d’ Afrique Tropicale) — to translate his comments would spoil them — ‘“‘comme c’était ma premiére visite 4 Kew il eut la délicate attention de venir me chercher a la gare de Victoria a Londres. Je vis un homme sérieux trés droit que me fit immédiatement penser a un officier anglais de l’armée britannique des Indes. Mais sous cet aspect un peu sévére je découvris rapidement un caractére particuliérement aimable.”’ Edgar was born near Frome in Somerset on 24 May 1906. His father George Bertram Milne- Redhead was a keen gardener and his grandfather Richard, a great traveller, seed collector and Fellow of the Linnean Society, had established a fine garden at Holden Clough near Clitheroe. His mother Agnes was interested in classical music and croquet and after the death of Edgar’s elder sister Rosamund at only 16 developed a strongly protective manner towards him. Edgar’s early schooling was at The Old Ryde preparatory school in Bournemouth. In 1920 the family moved to Cheltenham enabling him to attend the college as a day boy and to enjoy the fine countryside. He went up to Gonville and Caius College, Cambridge in 1925 and read Natural Sciences, taking a particular interest in botany after meeting the legendary Humphrey Gilbert-Carter and gaining a half-blue for rifle shooting. Examinations, however, were not his forte and rather than take Part II of the Tripos he applied for a post at Kew. No posts being available he accepted an unpaid position for several months. A terse announcement in the Bulletin of miscellaneous information, Kew records that in 1929 C. E. Hubbard and E. Milne-Redhead had been appointed Temporary Subassistants in the Herbarium — there were no flattering titles for the lower ranks in those days! For a year he worked successively on plants of Europe, Canada, Fiji and elsewhere; then a remarkable opportunity occurred in March 1930. The then Director of Kew, Sir Arthur Hill, was asked by the Colonial Office to suggest a botanist to assist with an aerial survey of what is now Zambia and offered to second Edgar to the scheme. The offer was eagerly accepted and he spent four and half months in Mwinilunga District when he prepared some of the most elegant herbarium specimens ever to have come out of Africa. His unpublished report on the interpretation of vegetation by aerial surveys is in the library at Kew. In 1933 he married Olive, the sister of a senior colleague Kenneth Airy Shaw. She became an excellent botanical artist and her drawings of African plants grace the work of many botanists (including one of my earliest papers). Their golden wedding was celebrated with friends at Great Horkesley. She survived him until September 1997. For some years the Empire Marketing Board had funded several posts at Kew including Edgar’s but in 1935 the Board came to an end and Edgar and others joined the official Kew staff. In 1936 he succeeded John Hutchinson as head of the tropical African Section, a position he held until 1959. Leading a University Travel Club expedition to the Austrian Tyrol he gained some insight into the European mountain flora. A return to Mwinilunga was possible in 1937, thanks mostly to the hospitality of his friends Capt. and Mrs K. R. Patterson but after four and half months his request to stay longer in order to encompass the second half of the rainy season was refused for no obvious reason. Edgar was embittered by this ill-considered decision typical of the attitude of half-witted administrators to scientists everywhere. The second collection was even better than the first and it is a great pity it was not written up as a whole. Nevertheless many of the more striking new species Edgar were described in a series of contributions to Hooker’s Icones Plantarum and in a series entitled African Plants in the Bulletin of miscellaneous information, Kew. The collection was kept in the basement and often missed by visiting researchers, in fact frequently hidden from them; only people judged to be good workers were given access. In my opinion it would have been preferable if the whole collection had been named up even if only to genus and the duplicates distributed. In the end decades were to pass before it was all dealt with and accessible to everyone. This collection has been one of the main sources of information for Mwinilunga District for Flora Zambesiaca. Before the war Edgar gave up quite a lot of time (unlike most of his colleagues) encouraging the student gardeners by giving talks and taking them out on botanical excursions. Work at Kew was of course totally disrupted by the war. Edgar who had been commissioned as a Second Lieutenant in the Territorial Army in 1929 and served for ten years with the 30th (Surrey) Searchlight Battalion, Royal Engineers was called up when the Air Defence services were mobilised in August 1939. He became a gunner in 1940 when the searchlight units were transferred to the Royal Artillery and in November were drafted to West Africa and attached to the Royal West 130 OBITUARIES African Frontier Force. Edgar rose from the rank of Captain to Temporary Major and managed to collect a few plants and make some observations in Nigeria, Sierra Leone and the Gold Coast (now Ghana). On his return to England in early 1942 he became a Sector Searchlight Control Officer working with R.A.F. Fighter Command stationed at the R.A.F. Radar Station at East Hill near Houghton Regis on night interception of German bombers. Daylight hours were spent exploring on a bicycle around Dunstable and the nearby chalk hill flora. He also collected Hemiptera- Heteroptera, at that time very poorly known in Bedfordshire, and the results were written up by Airy Shaw. He soon made friends with local naturalists, particularly John Dony and Vic Chambers and also with two youngsters, myself and Peter Taylor who both owe him a great deal. Peter was later to leave his engineering job to join him at Kew and Edgar was instrumental in obtaining a job for me with Peter Greenway in what was then Tanganyika. At the request of James Fisher, Edgar undertook the rook census for South Bedfordshire in 1944. Luton Hoo was rich in rookeries but was the local testing ground for tanks by Commer Cars in Luton. It was definitely off limits to everyone. Edgar decided that if he wrote to the War Office for permission, by the time it had arrived the rooks would have left so he put on his uniform and walked smartly in, receiving a salute from the sentry who did not ask him his business. Once in he was free to wander all over the area counting nests, with tanks charging about the drives and tracks. He completed the task in about two hours and walked out past the sentry with no questions asked. It resulted in a very big count, more nests per hectare than anywhere else in his allotted area. When he returned to Kew after the war he soon gained the rank of Principal Scientific Officer and set out to build up the African section. He encouraged many local amateur collectors (mostly but not all colonial government officers) in Tropical Africa to send in material and most responded to his request to collect only high quality material with meticulous notes. His great success is immediately apparent to anyone working at Kew who has to compare African plants with those from other areas. His standard slowly spread to all collectors on expeditions from Kew. Nothing annoyed him more than a poor specimen with scrappy notes. The Flora of West Tropical Africa had long been completed and in fact work on a new edition was shortly to begin. A Flora for the other side of the continent was mooted some years before the war and Edgar was instrumental in 1949 in initiating the Colonial Office programme for such a project with provisions built in for a number of major expeditions. This vast undertaking was supposed to be finished in 20 years but is still far from completion. Edgar was the main editor until his retirement. His first co-editor was the Keeper of the Herbarium, W. B. Turrill, who thought a species could be written up in 20 minutes whereas a week is nearer the mark. Hubbard followed Turrill and in 1965 Roger Polhill, who continued alone after Edgar’s retirement. . It is equally true of both professional and amateur botanists that they are associations of friends and the founding of A.E.T.F.A.T. in 1949 by Edgar, together with Arthur Exell of the British Museum and Jean Léonard of Brussels, was from the first an informal organisation of friends which has been of immense importance to students of African botany. It still flourishes and membership has grown over twenty fold but the informality of the small band of original friends still to some extent prevails. In 1956 Edgar and Peter Taylor undertook an eight month collecting expedition to East Africa spending most of their time in Songea District in southernmost Tanzania, an area where only a very few small collections had been made. Their collaboration resulted in 5000 quite perfectly prepared gatherings of plants, mostly with many duplicates, which added immensely to the knowledge of East African plants. In 1959 Edgar became Deputy Keeper of the Herbarium and editor of the Kew bulletin (the channel for publication of most scientific work at Kew), posts he retained until his retirement in 1971. He was very disappointed not to get the post of Keeper (for which he was ideally suited) — it went instead to C. E. Hubbard who had come into scientific work from the gardens and was academically unqualified. One must not forget of course that Hubbard had become one of the world’s foremost grass experts and was much better known worldwide than Edgar and to reward him with such a post was only just, but Hubbard was a poor administrator. Edgar worked well at his new posts and despite his authoritarian manner was intensely loyal to his staff. Edgar had long been deeply involved in various aspects of conservation, being an Associate of the Royal Society for Nature Conservation from 1948 and on the Standing Committee of “The Countryside in 1970” the third of a series of conferences, supported by the Duke of Edinburgh and OBITUARIES 131 the then Prime Minister Edward Heath, to assess land use and environmental responsibilities. One of his last successes at Kew was to persuade the new Director, Prof. Heslop-Harrison, to set up a conservation unit at Kew. This was achieved in 1972 in time for Kew to participate in the first meeting of the Convention for International Trade in Endangered Species (CITES) (in Washington 1973). Kew has been concerned with conservation ever since. The retirement dinner for Edgar was a very well attended affair with hand-painted menus by Pat Halliday. I think he was pleased to be going and looking forward to having time for other interests, which was fortunately vouchsafed him. He had of course always been deeply interested in British botany and participated fully in the mapping schemes which resulted in the Atlas of the British flora in 1962. He joined the B.S.B.I. in 1929 and was on the Council as early as 1939 and at various other times, 1946-1950, 1951-1955 and again in 1957. He served on the Development Committee (later Development and Rules) from 1947-1967 being secretary in 1950, also on the Field Work Committee (from 1947), Maps Committee (1952-1965), Records Committee (1966-1967) and Conservation Committee (1952- 1985), serving as its Chairman for over 10 years. This committee was very involved with the “Conservation of wild creatures and wild plants” Bill completed in 1975. Edgar represented the B.S.B.1. on the Council for Nature (1960-1969) of which he was a founder member and on the Wild Plant Protection Working Party in 1965 formed to promote legislation for wild plants in Britain. The publication and wide distribution to the public of the Code of Conduct for the Conservation of Wild Plants and the ‘“‘Save our endangered wild flowers” poster were mainly due to him. He became President in 1970-1971 and was eventually awarded with honorary membership; at the end he was the Society’s oldest living member. In 1964-1967 he was a leading campaigner on the Teesdale Appeal Committee to save the wonderful relict flora of Cow Green from being destroyed by a reservoir, and representing B.S.B.I. he founded the Cypripedium Committee in 1970 to look after the single remaining specimen of the Lady’s slipper orchid. Edgar used to tell how he persuaded the rather suspicious local naturalists to a first meeting held in a Grassington pub after a good lunch. The policy agreed differed from that of the locals who had previously believed in total secrecy, but now with co-operation and wider resources available the single remaining plant was saved, the seed collected and after many years seedlings successfully raised and now planted out at the original and new sites. Edgar also helped organise the annual scrub clearance on the Goring Scarp to preserve other scheduled orchids. He also encouraged one of the first County Wildlife Roadside Verge Schemes. He was also instrumental in setting up the smallest nature reserve in the world to conserve Ranunculus ophioglossifolius in Gloucestershire at Badgworth, designated an S.S.S.I. in 1949. It was leased by the Society for the Promotion of Nature Reserves to the Gloucestershire Trust for Nature Conservation in 1962 when Edgar was Chairman of the Management Committee of that Trust (of which he was a founder member). Its success was due to him noticing that the species responded well to trampling and recommending that cattle should be allowed to disturb the ground. His well known survey of the native black poplar occupied him for many years and was his chief interest — in fact he was identifying specimens until a few weeks before his death. His fondness for this tree he attributed to Gilbert-Carter (reports that Edgar searched for it as a schoolboy are incorrect but he and his father certainly made special records of it for the Flora of Gloucestershire). He published many reports on the progress of the survey which grew to national dimensions when it was taken up by The Daily Telegraph. His involvement with societies on various aspects of conservation was endless — he seemed to collect societies like some collect stamps (in fact he did that also). They include the Bedfordshire Natural History Society, B.B.O.N.T., Cotteswold Naturalists’ Field Club (Vice-President), Essex _ Naturalists’ Trust, Gloucestershire Naturalists’ Trust, The Kew Guild (President 1968-1969), Kew Lawn Tennis Club 1930-1970 (of which he was secretary for many years) (Olive was ladies’ champion on many occasions), Linnean Society (Vice-President 1953-54), London Natural History Society (Chairman of Nature Conservation Committee), his own local Nayland with Wissington Conservation Society (President), Norfolk Naturalists’ Trust, North Gloucestershire Naturalists’ Society (President), Ray Society (President, 1968-1971), Richmond Society, Société Royale de Botanique de Belgique (Honorary Member), Somerset Naturalists’ Trust, Suffolk Naturalists’ Trust (Hon. Vice-President), Surrey Naturalists’ Trust (Regional Representative for Western Boroughs of Greater London, Regional Secretary for Richmond, Representative on the County 132 OBITUARIES Naturalists’ Trust Committee of the Society for the Promotion of Nature Reserves, later the Royal Society for Nature Conservation and now the Wildlife Trust). Edgar certainly had his faults (who of any value does not), a short temper being one of them but it soon passed and he bore no grudge to its recipient, and to counter it his sense of humour was pronounced. He could be a most annoying editor both to authors and printers, suddenly deciding to alter something in page proof and not doing it properly so that a paper initially consistent in some recurring feature had some items half one way and half another; adding a last minute triviality could result in a real error and I and many authors were livid at the gratuitous mistakes in our work due to last minute fiddling, but he did bring new standards into the editing of the Kew bulletin. Little foibles could irritate — his habit of using the unit decimetre which was confusing to most people and his insistence that orbicular meant round and disk-like whereas I (and the Queen) used it for something spherical. He could be impractical; Peter Taylor was saddled with all the chores during their African trip and Edgar who was a poor linguist learnt no Swahili and was unable to communicate with their local staff. He did not write a great deal himself and was I believe temperamentally incapable of spending the immense time needed to revise a large group or write a long floristic account when no matter how long one spends errors and omissions are inevitable. Edgar was scathing about large works produced quickly by J. Hutchinson, W. Robyns and others, rather forgetting that one just had to have these practical works. He liked to attack more circumscribed problems thoroughly and excelled at producing published accounts of small groups — that on Montiniaceae for example is a typical gem. This sort of work gave full scope for his meticulous nature. This did not prevent him being interested in large groups. His long term work on Crotalaria and the Acanthaceae involved putting the herbarium material in order, sorting out the new species which needed describing but not actually doing much of the writing. Roger Polhill who joined Edgar at Kew in 1962 (following a collecting expedition arranged by Edgar in 1961 and after working with me in Nairobi for about a year) brought the revision of African Crotalaria (some 500 species) to fruition but would be the first to acknowledge Edgar’s considerable spade work. MR (as we all knew him) was awarded the Imperial Service Order in the 1967 New Year’s Honours List in recognition of his distinguished service at Kew and an M.B.E. in 1996 for his conservation work (it should have been more). Although he did not reach what a career pusher would consider any position of importance and is not mentioned in the recent history of Kew his work on Tropical African botany assures him of an honoured place in its history and 25 species have epithets commemorating him (list below). His beautifully prepared material will be valued by workers for as long as they last. His handwriting, exceedingly elegant and completely unmistakable will join those which when found on a label pronounce the determination as reliable. We at Kew are constantly and pleasantly reminded of our old friends since we use their determinations daily. His last months were spent at Nayland near his daughter Annette and her husband Basil Harley (well known for their natural history publishing) and later living with them at Martins, Gt Horkesley. Annette’s interest in entomology had been aroused by her uncle Kenneth Airy Shaw who was at Daglingworth (where much of Kew’s material was sent during the war) when she and her mother were at Cirencester during Edgar’s wartime absence. Edgar’s funeral at All Saints Church, Gt Horkesley was a joyous beautifully conducted celebration of a full life devoted to natural history. The church was decorated and his coffin strewn with branches from the Black Poplar he had planted at Martins. LIST OF PUBLICATIONS OF E. MILNE-REDHEAD Campanula orphanidea Boiss. Bulletin of miscellaneous information, Kew 1929: 142 (1929) (with W. B. Turrill) Variation in Anemone nemorosa. Journal of botany 70: 325-328 (1932) (with W. B. Turrill) Begonia plagioneura [sp. nov.]. Bulletin of miscellaneous information, Kew 1930: 269-271 (1930) Aporrhiza nitida Gilg in Tropical African plants: VII. Bulletin of miscellaneous information, Kew 1931: 272-273 (1931) The genus Strobilanthopsis. Bulletin of miscellaneous information, Kew 1932: 344-347 (1932) Campanulata spathulata §. et S. var. Giuseppii Milne-Redhead et Turrill [in Turrill, W. B. On the flora of the nearer east: XIII]. Bulletin of miscellaneous information, Kew 1932: 453 (1932) Tropical African plants: X. Bulletin of miscellaneous information, Kew 1933: 142-151 (1933). Contributions by OBITUARIES 133 E.M.R. Oxygonum fruticosum: 142; Acacia (Vulgares-Ataxacanthae) chariessa: 143-144; Albizzia (Eualbiz- zia-Microphyllae) struthiophylla: 144 Tropical African plants: XI. Bulletin of miscellaneous information, Kew 1933: 467-479 (1933). Contributions by E.M.R. Crotalaria (Simplicifoliae) vialis: 469; Disperma crenatum: 477; Aloé (Eualoé) ortholopha: 478-479, pl. 18 (with H. B. Christian) Hooker’s Icones Plantarum 33: t. 3213 Vigna nuda N.E.Br. t. 3214 Physostigma mesoponticum Taub. t. 3215 Oxygonum pachybasis Milne-Redhead t. 3216 Oxygonum tenerum Milne-Redhead (1933) Hooker’s Icones Plantarum 33: t. 3243 Crotalaria annua Milne-Redhead t. 3244 Crotalaria praecox Milne-Redhead t. 3245 Crotalaria streptorrhyncha Milne-Redhead t. 3246 Bolusia resupinata Milne-Redhead (1934) A buttercup. Gloucestershire countryside 1 (12): 187-188 (1934) The type of Ormocarpum kirkii. Bulletin of miscellaneous information, Kew 1934: 42-43 (1934) (with T. A. Sprague) Notes on the flora of Southern Africa V. The genus Ruellia in Thunberg’s Herbarium. Bulletin of miscellaneous information, Kew 1934: 264-265 (with R. A. Dyer) Tropical African plants: XII. Bulletin of miscellaneous information, Kew 1934: 301-307 (1934). Contributions by E.M.R. Albizzia sericocephala: 301; Vigna juncea major: 301-302; Disperma nudanthera: 304; Dyschoriste procumbens: 304; Monechma ciliatum: 304-305; Lapeyrousia schimperi: 307 New Combinations under Copaifera. Bulletin of miscellaneous information, Kew 1934: 400 (1934) Curtis’s botanical magazine 157: 9345 Hypericum kouytchense Léveillé 9349 Campanula propinqua var. grandiflora Milne-Redhead 9361 Verbena corymbosa Ruiz & Pavon 9373 Viburnum alnifolium Marshall (1934) Hooker’s Icones Plantarum 33: t. 3266 Blepharis hornbyae Milne-Redhead t. 3267 Anisotes umbrosus Milne-Redhead t. 3268 Anisotes bracteatus Milne-Redhead t. 3291 Barleria tetraglochin Milne-Redhead t. 3292 Barleria proxima Lindau t. 3293 Barleria quadrispina Lindau (1935) Tropical African plants: XIII. Bulletin of miscellaneous information, Kew 1935: 271-285 (1935). Contributions by E.M.R. Triplochiton zambesiacus: 271-272; Hibiscus mastersianus: 272-273; Crotalaria (Eucrotalaria) lotiformis: 274-276; Clausenopsis hildebrandtii: 278-279; Crossandra thomensis: 280-281; Disperma eremop- hilum: 282-283; Rhinacanthus pulcher: 283-284 Curtis’s botanical magazine 158: 9385 Choananthus cyrtanthiflorus (C.H. Wright) Rendle 9412 Rhodohypoxis baurii (Bak.) Nel (1935) Curtis’s botanical magazine 159: 9436 Campanula formanekiana Degan & Dorfler 9444 Adenophora morrisonensis Hayata Eranthemum of the “Flora of tropical Africa”. Bulletin of miscellaneous information, Kew 1936: 255-274 (1936) Tropical African plants: XIV. Bulletin of miscellaneous information, Kew 1936: 469-489 (1936). Contributions by E.M.R. Indigofera annua: 470; Indigofera shinyangensis: 471; Vigna macrorrhyncha: 473; Salix subserrata: 474; Fagaropsis hildebrandtii: 475; Turraea fischeri: 475; Anisotes dumosus: 487; Justicia salvioides: 488; Kniphofia (Laxiflorae) rogersii: 488; Anthericum suffruticosum: 489 Curtis’s botanical magazine 159: 9452 Sutera grandiflora (Galpin) Hiern (1936) Curtis’s botanical magazine 160: 9470 Chasmanthe caffra (Bak.) N.E.Br. 9487 Homeria collina (Thunb) Salisb. (1937) Tropical African plants: XVII. Bulletin of miscellaneous information, Kew 1937: 411-432 (1937). Contributions by E.M.R. Euphorbia eranthes (with R. A. Dyer): 413; Phyllanthus holostylus: 414; Dialium orientale: 415; Isoberlinia densiflora & magnistipulata: 415; Acacia eggelingii, hebeclada & nigrescens: 416-417; Desmodium setigerum: 417; Mundulea sericea: 417; Pleiotaxis arenaria: 424; Dicliptera arenaria: 427; Dicliptera capitata: 428; Dicliptera nemorum: 429; Justicia (Rostellularia) syncollotheca: 429; Monechma praecox: 430; Phaylopsis hispida: 431 134 OBITUARIES The genus Cordyla Loureiro. Repertorium Specierum novarum Regni vegetabilis 41: 227-235 (1937) Curtis’s botanical magazine 159: 9510 Homoglossum merianella (Thunb.) Bak. (1938) Choice between epithets of the same date. Bulletin of miscellaneous information, Kew 1939: 34-35 (1939) The correct name for Sesbania aculeata. Bulletin of miscellaneous information, Kew 1939: 159 (1939) (with T. A. Sprague) A new conception of the genus Ammocharis Herb. Journal of the Linnean Society (botany) 52 (342): 159-197 pl. 2, 3, 4 (1939) (with H. G. Schweickerdt) Hooker’s Icones Planatarum 34: t. 3388 Eriocaulon strictum Milne-Redhead t. 3389 Eriocaulon annuum Milne-Redhead t. 3390 Gerardina eylesiana Milne-Redhead (1939) Curtis’s botanical magazine 161: 9556 Campanula incurva Auch. ex DC. (1939) Curtis’s botanical magazine 162: 9568 Campanula affinis Roem. & Schultes (1939) Tropical African plants: XVIII. Bulletin of miscellaneous information, Kew 1940: 49-66 (1940). Contributions by E.M.R. Crotalaria diloloensis: 52; Crotalaria elisabethae: 52; Hypoestes rosea: 64; Barleria (Somalia) phaylopsis: 65 Hooker’s Icones Plantarum 35: t. 3417 Buchnera prorepens Engl. & Gilg (1940) Curtis’s botanical magazine 164: 9676 Erica pillansii Bolus (1946) Hooker’s Icones Plantarum 35: t. 3459 Fadogia spectabilis Milne-Redhead t. 3460 Piliostigma thonningii (Hochst.) Milne-Redhead (1947) Tropical African plants: XIX. Kew bulletin 2: 23-35 (1947). Contributions by E.M.R. Gomphrena celosioides: 23; Impatiens niamniamensis: 23; Syzygium huillense: 24; Baphia pyrifolia: 26; Crotalaria unicaulis: 26; Vigna paludosa: 27; Morinda morindoides: 31; Pavetta coriacea: 31; Pentas zanzibarica: 31; Aloe zanzibarica: 33; Anthericum dimorphum: 33; Crinum minimum: 33; Juncus dregeanus: 35 A probably [sic] natural hybrid in Hymenocardia Tul. Kew bulletin 2: 46 (1947) Cerastium brachypetalum Pers: in Britain. The naturalist 1947 [822]: 95—96 (1947) Ferdinandia Welw. ex Seem [Bignoniaceae], an unintentional orthographic error. Kew bulletin 3: 170-171 (1948) Tropical African plants: XX. Kew bulletin 3: 449-473 (1949). Contributions by E.M. R. Cladostemon kirkii: 449- 450; Bergia capensis: 450; Polycarpon prostratum: 451-452; Robbairea delileana: 452; Philoxerus vermicular- is: 452-453; Geranium ocellaturn var. sublaeve: 453; Geranium yemense: 453-454; Hypericum oligandrum: 454; Hypericum stolzii: 455-456; Corchorus hochstetteri: 456; Oldfieldia somalensis: 456; Euphorbia geniculata: 457-458; Vigna juncea: 458; Nauclea latifolia: 459; Ceropegia yorubana: 465; Chrysanthellum americanum: 466; Solanum bifurcatum: 467; Dopatrium luteum: 467; Crocosmia pauciflora: 469; Urginea glaucescens: 470-471; Paepalanthus lamarckii: 472 The positions of Kew in relation to taxonomic and ecological researches in Africa. Communication No. B (h) 22 at African Regional Scientific Conference, Johannesburg 4 pp. Cae Hooker’s Icones Plantarum 35: t. 3477 Loranthus dependens Engl. t. 3478 Petamenes vaginifer Milne-Redhead t. 3479 A et B Ensete homblei (Bequaert ex DeWild.) E. E. Cheesm. t. 3480 Euphorbia erythrocephala Bally et Milne-Redhead t. 3481 Euphorbia asclepiadea Milne-Redhead (1950) Curtis’s botanical magazine 167: 111 Crotalaria raffillii Milne-Redhead (1950) Clef pratique des Marantacées Congolaises. Bulletin de la Société royale de Botanique de Belgique 83: 5—32 (1950) (with J. Léonard & W. Mullenders) Notes on African Marantaceae: 1. Kew bulletin §: 157-163 (1950) Variation in leaf-shape within a species: some examples from the Gold Coast. Kew bulletin 5: 261 (1950) Tropical African plants: XXI. Kew bulletin §: 335-384 (1951). Contributions by E.M.R. Spergula fallax: 338; Robbairea delileana: 340; Geranium purpureum: 341; Hypericum scioanum & humbertii: 343-344 (with J. B. Gillett); Pterygota macrocarpa: 348; Adenocline acuta: 349; Crotalaria grandistipulata: 349-351; Neorautane- nia pseudopachyrrhiza: 355; Eriosema burkei, cryptanthum, velutinum, molle & andohii: 355-363; Lasiodis- cus zenkeri: 366; Emilia praetermissa: 375-376; Crassocephalum bauchiense, montuosum, baoulense & mannii: 376-377; Exochaenium pygmaeum: 377-378; Cyphia erecta: 378-379; Englerastrum gracillimum: 380; Monechma ciliatum: 381; Vellozia schnitzleinia var. occidentalis: 381-382; Amorphophallus abyssinicus: 382- 384 Tropical African plants: XXII. Ranunculaceae. Kew bulletin 5: 389 (1951) (with W. B. Turrill) A new species of Ranunculus from East Africa. Kew bulletin 6: 147 (1951) (with W. B. Turrill) OBITUARIES 135 The identity of Polygala phylicoides Thunb. Kew bulletin 6: 148 (1951) Les espéces africaines du genre Clematopsis Boj. ex Hutch. Bulletin de la Société royale de Botanique de Belgique 83: 407-427 (1951) (with A. W. Exell & J. Léonard) Field Meeting to Taunton. B.S.B.J. Year Book 1951: 37-39 (1951) Field Meeting to Isle of Man. B.S.B.I. Year Book 1952: 32-34 (1952) Notes on African Marantaceae. Kew bulletin 7: 167-170 (1952) Flora of Tropical East Africa. Ranunculaceae, pp. 23 (1952) (with W. B. Turrill) Flora of Tropical East Africa. Marantaceae, pp. 11 (1952) The vegetation of the Solwezi and Mwinilunga Districts of Northern Rhodesia. Symposium 1 of A.E.T.F.A.T., Flora and vegetation of Tropical Africa: 125 (1953) (reprinted from Lejeunia 16) Lepidagathis collina in Tropical African plants XXIII. Kew bulletin 8: 119 (1953) Tropical African plants XXIV. Kew bulletin 8: 431-445 (1953). Contributions by E.M.R. Ranunculus oligocarpus: 431; Hypericum keniense & annulatum: 434-437; Vismia pauciflora: 437; Justicia praecox: 444 Flora of Tropical East Africa. Hypericaceae, p. 23 (1953) Zornia in tropical Africa. Boletim da sociedade broteriana 28: 79-104 (1954) Distributional ranges of flowering plants in tropical Africa. Proceedings of the Linnean Society 165: 25—35 (1954) Hooker’s Icones Plantarum 36: t. 3541 A & B Montinia caryophyllacea Thunb. t. 3542 Montinia caryophyllacea Thunb. t. 3543 Grevea eggelingii Milne-Redhead t. 3544 Grevea eggelingii Milne-Redhead [pp. 1-16] Montiniaceae (1956) Flora of Tropical East Africa. Memérias da Sociedade broteriana 13: 57-59 (1958) Ranunculaceae in Flora Zambesiaca 1: 89-102 (1960) (with A. W. Exell) Miscellaneous notes on African species of Crotalaria: 1. Kew bulletin 15: 157-167 (1961) The typification of Hedysarum diphyllum L. Kew bulletin 17: 73-74 (1963) (with J. E. Dandy) Dioscorea (Borderea) gillettii. Contribution to Tropical African plants XX VII. Kew bulletin 17: 177-179 (1963) B.S.B.I. Conference Reports No. 8. The conservation of the British flora, pp. 90, 1963 [editor] Birds in Richmond during the Winter of 1962-63. The Surrey naturalist. Annual Report. April 1965: 25-27 (?1965) Curtis’s botanical magazine 175: 474 Arisaema schimperianum Schott (1965) Progress of the Flora of Tropical East Africa (Proceedings of 5th meeting of A.E.T.F.A.T.). Webbia 19: 891- 892 (1965) The firearms act and Conservation. The Surrey naturalist. Annual Report for 1966: 41-42 (1967) Flora of Tropical East Africa (Proceedings of 6th meeting of A.E.T.F.A.T.). Acta phytogeographica suecica 54: 291 (1968) The London Nature Conservation Committee. The Surrey naturalist. Annual Report for 1969: 36-38 (1970) Progress of the Flora of Tropical East Africa (Proceedings of 7th meeting of A.E.T.F.A.T.) Mitteilungen aus der botanischen Staatssammlung Miinchen 10: 66-67 (1971) Leguminosae — Zornia. Contribution to Tropical African plants XXXI. Kew bulletin 25: 178-180 (1971) Presidential address 1970. Botanical conservation in Britain, past, present and future. Watsonia 8: 195-203 CI972) Notes on African Dioscorea. Kew bulletin 26: 573-576 (1972) Flora of Tropical East Africa. Leguminosae-Papilionoideae (1971). Contributed Zornia, pp. 442-450; Gamwellia, Rothia & Lotononis, pp. 808-817; Adenocarpus, pp. 1009-1011 Ranunculaceae in Flora de Mocambique 4, p. 16 (1973) (with A. W. Exell & M. L. Gongalves) Orchis militaris. B.S.B.I. news 3 (1): 11- 12 (1974) Reserves for arable weeds. B.S.B.I. news 3 (2): 10 (1974) Obituary. Dr H. Milne-Redhead. B.S.B.I. news 3 (2): 15-16 (1974) Letter to Mr Beckett. B.S.B.I. news 3 (2): 21 (1974) (with F. H. Perring) Report. B.S.B.I. Black Poplar survey. Watsonia 10: 299 (1974) Obituary. Humphrey Milne-Redhead (1906-1974). Watsonia 10: 449-450 (1975) Flora of Tropical East Africa. Dioscoreaceae, p. 25 (1975) Pholeus [sic] phalangioides at Walberswick [Pholcus]. Suffolk natural history 17: 67 (1975) Black Poplar survey. Watsonia 10: 295-296 (1975) Rex Graham Reserve. B.S.B.I. news 9: 9 (1975) Black Poplar survey. B.S.B.I. news 9: 10-12 (1975) A remarkable population of Oper vulgatum L. in Suffolk. Watsonia 10: 415—416 (1975) (with P. J. O. Trist) Another Suffolk record for Balea perversa (L.). Suffolk natural history 17: 66-67 (1975) Rex Graham Reserve. B.S.B.J. news 14: 8-9 (1976) Obituary. Victor Samuel Summerhayes (1887-1974). Watsonia 11: 90-92 (1976) Report. B.S.B.I. Black Poplar survey. Watsonia 11: 180-181 (1976) 136 OBITUARIES Populus nigra Linn. — a rare Essex tree. Essex naturalist 33: 155-156 (1976) A cosmopolitan psocid in Suffolk. Suffolk natural history 17: 141 (1976) The Black Poplar survey — a progress report. B.S.B.I. news 15: 9 (1977) More about the Black Poplar survey. B.S.B.I. news 16: 13-14 (1977) An appreciation from an absentee member. Bedfordshire Natural History Society newsletter 26: 1 (1977) Chelostoma campanularum Kirby, Hymenoptera. Suffolk natural history 17: 379-380 (1978) Obituary. Mary Alice Eleanor Richards (1885-1977). Watsonia 12: 187-190 (1978) Field Meeting Report. Tewkesbury, Glouc. May 8th. Watsonia 12: 66 (1978) Lapsana intermedia. B.S.B.I. news 20: 24 (1978) John Dony at 80. B.S.B.I. news 22: 6 (1979) An under-recorded millipede from Suffolk. Suffolk natural history 18: 148 (1980) Orchids at Waltham Abbey — Essex. B.S. B.I. news 25: 22-23 (1980) Notes on some Suffolk sawflies. Suffolk natural history 18: 293-295 (1982) The Black Poplar survey. B.S.B.J. news 33: 6 (1983) [errata in 34: 38 (1983)] Autumn Lady’s Tresses. B.S. B.I. news 36: 27 (1984) In pursuit of the poplar. Natural world 10: 26-28 (1984) Obituary. Maybud [May] Sherwood Campbell (1903-1982). Watsonia 15: 157-160 (1984) (correction: 419 (1985)) The not so hardy British Oak. B.S.B.I. news 39: 19 (1985) Where my caravan has rested. B.S.B.I. news 40: 8 (1985) More hosts of Orobanche. B.S.B.I. news 41: 29 (1985) Field Meeting Report. Vale of Aylesbury, Buckinghamshire, 26th-27th May 1984. Watsonia 16: 113 (1986) Cotton tree in Lancashire. B.S.B.I. news 45: 32 (1987) Conservation of Muscari atlanticum Boiss. & Reuter in Suffolk. B.S.B.I. news 47: 32-33 (1987) An uncommon orchid hybrid, Dactylorhiza X transiens. Suffolk natural history 24: 82-83 (1988) An unusual gall. Suffolk natural history 24: 100 (1988) Where are Suffolk’s earwigs? Suffolk natural history 25: 44 (1989) The B.S.B.I. Black Poplar survey, 1973-88. Watsonia 18: 1-5 (1990) More about white flowers. B.S.B.I. news 58: 12-13 (1991) Cutting roadside verges. B.S.B.I. news 60: 31 (1992) Mega-Mustard (Sinapis alba L.). B.S.B.I. news 60: 65 (1992) Protected roadside verges. White Admiral 22: 13 (1992) B.S.B.I. Black Poplar survey, 1973-8. White Admiral 25: 35 (1993) Reminiscences of a naturalist nearly 50 years ago. The Muntjac 90: 1 (1993) Edgar reflects: His first African expedition in 1930. White Admiral 26: 8-9 (1993) Edgar Milne-Redhead — a correction. White Admiral 27: 39 (1994) Obituary. Sonia C. Holland (1912-1993). Watsonia 20: 171-173 (1994) Viola pensylvanica — a mystery solved? Transactions of the Suffolk Naturalists’ Society 28: 59 (1992) B.S.B.I. Black Poplar survey. B.S.B.I. news 67: 22-23 (1994) Black Poplar hunt. White Admiral 29: 27-28 (1994) Ticking them off. White Admiral 30: 38 (1995) Tragopogon in Suffolk. B.S.B.I. news 70: 21 (1995) Recollections of the early years 1. The Muntjac 103: 3 (1996) He also made numerous contributions particularly of birds to various parts of “Additions to the wild fauna and flora of the Royal Botanic Gardens Kew” published in the Kew bulletin. Eponymous epithets Aloe milne-redheadii Christian, Amphiasma redheadii Brem., Ascolepis erythrocephala S. S. Hooper, Cissus milnei Verdc., Clerodendrum milne-redheadii Moldenke, Crotalaria milneana Wilczek, Digitaria redheadii (C. E. Hubbard) Clayton, Helichrysum milne-redheadii Brenan, Hypericum milne-redheadii Gilli, Indigofera milne- redheadii J. B. Gillett, Ipomoea milnei Verdc., Lantana milne-redheadii Moldenke, Lobelia milneana E. Wimmer, Monotes redheadii Duvign., Nymphoides milnei A. Raynal, Pandiaka milnei Suesseng & Overk., Pavetta redheadii Bremek., Polystachya erythrocephala Summerhayes, Rotala milne-redheadii A. Fernandes & Diniz, Spermacoce milnei Verdc., Strychnos milne-redheadii Duvign. & Staquet, Stylochaeton milneanus Mayo, Triglochin milnei H. Horn, Vernonia milne-redheadii H. Wild. There is also a Euphorbia erythrocephala Bally & _ Milne-Redhead which has red cyathia but was probably meant as a bit of a joke! OBITUARIES ey) ACKNOWLEDGMENTS I am grateful to the writers of several previous obituaries, to Mary Briggs for information on British connections, to Annette and Basil Harley for vetting the manuscript, to Rosemary Davies for the list of eponyms many of which I would otherwise have missed and to Suzy Dickerson for some of his publications. E.M.R. kept no list and I have doubtless missed a great many. B. VERDCOURT JOHN GRANT ROGER (1909—1997) Grant Roger was an Aberdonian whose youthful passion for botany led him first to the tropics as a tea planter on Sumatra. Disillusioned with colonial ways, he returned to Aberdeen University to study botany under Professor J. R. Matthews. After graduating with distinction, he joined the Regional Museum in Aberdeen, but in 1944 became Assistant Keeper in charge of Botany in the Manchester Museum. When the Nature Conservancy was set up in 1949, a career in wildlife conservation beckoned, and the next year he joined the Edinburgh headquarters staff, to participate in developing the Scottish programme. Grant’s botanical expertise led to field survey and assessment widely over Scotland, developing a valuable network of contacts and spreading the message as an ambassador for conservation. His later role as Regional Officer for north-east Scotland gave him jurisdiction over the magnificent Cairngorms and other parts of his home territory. He helped to set up the National Nature Reserves over the Cairngorms, and other famous areas, including Glens Clova and Caenlochan, Muir of Dinnet, Sands of Forvie, St Cyrus and the Morrone Birchwood at Braemar, his spiritual home. These reserves and the more numerous Sites of Special Scientific Interest were his most enduring achievement in conservation, and their management and monitoring became a major part of his work. The administrative side was a necessary but less appealing part of the job, and Grant was happiest in the field, keeping an eye on reality in nature and discussing the flora and fauna with Wardens and other people on the ground. He was a brilliant photographer, and formed a unique pictorial record of his field experiences, which he drew upon in his many and much acclaimed lectures about botanical conservation. Grant knew the Scottish flora inside out, and especially the mountain plants, but had a wide knowledge of British botany and its history besides. His own searches expanded knowledge of plant distribution considerably. A revered figure in the Botanical Society of Edinburgh (now Scotland), he became its President from 1966-68. After retirement from the Nature Conservancy in 1970, he was honoured with the Queen’s Jubilee Medal for his work on wildlife conservation in Scotland. For long afterwards, he gave annual courses in field botany at Kondrogan Field Studies Centre, where students continued to absorb his knowledge and wisdom. Amongst those who knew him, it is Grant Roger the person who is remembered above all. He had an endearing modesty, gentleness and old-world courtesy, and a total lack of conceit or pretence, yet a keen sense of fun, that won him a great many friends. A delightful companion, his priceless anecdotes, aphorisms and wry comments on the human condition, enlivened his learned discourses on matters botanical. His heroes were the self-taught Victorian naturalists of humble origin such as Russel Wallace, Richard Spruce and Hugh Miller whose selfless dedication to the pursuit of knowledge he so much admired. His wide interests were to be seen in his home full of books, covering not only natural history, but also art, music, literature, philosophy, history, castles, heraldry, ships and railways. Grant was blessed with a long and happy married life of 55 years, to Jean, who supported him devotedly throughout, including nursing him through a final difficult period of heart problems. Their warm hospitality to their many friends and visitors was legendary. Home life with Jean and children Neil, Lindsay, David and John meant a great deal to Grant, and annual family holidays in interesting places were a great event. Their lives were, sadly, clouded by John’s tragic death in a 138 OBITUARIES climbing accident in 1975. The company of his grandsons Ben and Sam became a great delight to Grant in later life. Grant Roger was a lovable person, whose integrity, high ideals and generosity of spirit will remain an abiding memory and inspiration to those fortunate to have known him. D. A. RATCLIFFE Watsonia 22: 139-142 (1998) 139 Report THE ANNUAL GENERAL MEETING, 17 MAY 1997 The Annual General Meeting of the Society was held at the Dorset County Museum, High West Street, Dorchester, by kind invitation of the Curator, Mr Richard de Peyer, at 11.00 a.m. The President, Mr D. A. Pearman took the Chair in the presence of 112 members. Apologies for absence were read and Minutes of the 1996 Annual General Meeting, published in Watsonia 21: 301-303 (1997), were accepted as correct, approved and signed by the President. REPORT OF COUNCIL The President took members through the Report of the Council, which had been previously circulated to members, commenting on the main achievements of the Society during the year, including the progress of Atlas 2000 and the computerisation of vice-county recorders. Mr R. G. Ellis was congratulated on producing his first Annual Report and Mr M. Walpole on his 26th and final Treasurer’s Report and Accounts. The adoption of both Reports was proposed by R. M. Burton, seconded by R. M. Walls and accepted unanimously. RULE CHANGES The President explained the reasoning behind the proposed amendment to Rule 9 where Council had proposed a change from: “The President shall be elected to serve for a term of two years and shall not be eligible for immediate re-election. . .” to ‘“The President shall be elected to serve for a term of two or three years, according to the wishes of Council and the nominee on taking office, and shall not then be eligible for immediate re-election . . .”. After some discussion this amendment was proposed by C. A. Stace, seconded by A. O. Chater and carried with one opposing vote and some abstentions. As a result of this rule change, R. G. Ellis proposed and T. G. Evans seconded that D. A. Pearman be re-elected President for a further year and this was carried unanimously. In introducing the proposed amendment to Rule 22, the President mentioned the outstanding contribution Messrs Grant Thornton had made to the successful running of the Society in auditing our accounts for the past 25 years. The proposal by Council to change Rule 22 was to give the Society flexibility when selecting a replacement for Messrs Grant Thornton. Council had proposed a change from: “‘An auditor shall be appointed by the members at the Annual General Meeting and the accounts, having been audited by him, shall be approved by the Council before presentation to the next Annual General Meeting.” to “‘An Auditor or Independent Examiner shall be appointed by the members at the Annual General Meeting to report on the Financial Statements and Accounts which shall then be approved by Council before presentation to the next Annual General Meeting.” The amendment was proposed by D. E. Allen, seconded by M. Walpole and carried unanimously. ELECTION OF VICE-PRESIDENT The President commented on the importance to the Society of having experienced members as Vice-Presidents who could assist the Chair when necessary. He paid tribute to Mr M. Walpole who ~ had done so much for the Society as Hon. Treasurer and Chairman of Publications Committee for over a quarter of acentury. Mrs M. Briggs then gave an appreciation of Mr Walpole with whom she had worked so closely as Hon. General Secretary for almost the same quarter of a century. The election of Mr M. Walpole was proposed by the President, seconded by Dr S. L. Jury and carried 140 REPORT with acclamation. Mr Walpole then gave a short address in which he expressed his gratitude for being allowed to serve the Society for so many years. ELECTION OF HON. GENERAL SECRETARY Council had nominated Mr R. G. Ellis. His election was proposed by Mr T. G. Evans, seconded by Mr R. G. Woods and carried unanimously. ELECTION OF HON. TREASURER With the resignation of Mr M. Walpole there was a vacancy for the post of Hon. Treasurer. Council had nominated Mr M. E. Braithwaite. His election was proposed by Mr A. O. Chater, seconded by Miss E. Young and carried unanimously. The President then warmly thanked the Editors of Watsonia and B.S.B.I. news, the compiler of B.S.B.I. abstracts, and all representatives on Council and other Committees for the hard work they carried out, voluntarily, on behalf of the Society. This was greeted by applause. ELECTION OF COUNCIL MEMBERS In accordance with Rule 11 nominations had been received for Mr C. R. Boon, Prof. M. Crawley and Dr S. J. Whild. Profiles had been published with proposers and seconders, and election of these members was approved unanimously. ELECTION OF HONORARY MEMBERS The President mentioned that there was a slight break with tradition in that no fewer than five members had been nominated for Honorary Membership of the Society this year. All had made outstanding contributions to the Society and to field botany in general. Sponsors for all five gave short appreciations of their respective candidates (to be published in B.S.B.I. news) and their election was carried unanimously with warm applause. The five new Honorary Members (and sponsors) were: Mr A. C. Jermy (Mr D. A. Pearman), Mr J. Ounsted (Miss A. Burns), Dr F. H. Perring (Mr M. Walpole), Dr F. Rose (Mr D. Streeter) and Mr P. D. Sell (Dr C. D. Preston & Dr S. M. Walters). PRESIDENT’S AWARD David Bellamy (President Wild Flower Society) and David Pearman (President B.S.B.I.) had both unhesitatingly recommended Richard Mabey for his Flora Britannica. Both Presidents thought this book quite superb and it had opened up botany to many people who would not otherwise have had an interest in the subject. ELECTION OF HONORARY AUDITOR OR INDEPENDENT EXAMINER The President again warmly thanked Grant Thornton, West Walk, Leicester for their exemplary auditing of our accounts. Their decision not to seek re-election resulted in a vacancy. The new Hon. Treasurer Mr M. E. Braithwaite had approached one of his local ‘“‘competitors”, Mr John Coates, of Greaves West & Ayre, Chartered Accountants, Berwick upon Tweed, who had agreed to stand for REPORT 141 election. He was proposed by Mr D. A. Pearman, seconded by Mr J. Ounsted and carried unanimously. There being no other business, the meeting closed at 12.40 p.m. GwyNnNn ELLIS FIELD EXCURSIONS HELD IN CONJUNCTION WITH THE A.G.M. Four field meetings were held over the weekend; all were very well attended, and we have to thank our leaders, Dr Sue Eden, Miss Anne Horsfall and Dr Humphrey Bowen for their sterling work. FRIDAY 16 MAY, KINGCOMBE This was meant to be a foretaste for early arrivals, yet over 50 members assembled for the walk, which was by courtesy of the Dorset Wildlife Trust, who own and farm the 158 ha reserve. Kingcombe is certainly one of the largest blocks of unimproved neutral grassland in southern England, and has very good numbers of plants, once common, but now rare outside reserves. The party saw Genista anglica (Petty Whin), Pedicularis sylvatica (Lousewort), Ophioglossum vulgatum (Adder’s-tongue), the leaves of Serratula tinctoria (Saw-wort), Succisa pratensis (Devil’s-bit Scabious) and Oenanthe pimpinelloides (Corky-fruited Water-dropwort), and a huge quantity of sedges, of which over 20 species occur here. There are good populations of Marsh and Small Pear!- bordered Fritillaries too. The highlight of the meeting was the discovery of a small patch of Botrychium lunaria (Moonwort) which, in Dorset, has only been seen once in the last 20 years. Sixty three members then found their way through the lanes to Frome St Quintin and tea and obligatory garden tours. SATURDAY 17 MAY, WINFRITH HEATH A coach load and some private cars met here for an early season heathland visit. The best area was a closely rabbit-grazed “‘village green”, where Cerastium semidecandrum (Little Mouse-ear), Crassula tillaea (Mossy Stonecrop) (very frequent), Ranunculus parviflorus (Small-flowered Buttercup), Trifolium micranthum (Slender Trefoil), T. ornithopodioides (Bird’s-foot Clover), T. scabrum (Rough Clover), T. striatum (Knotted Clover) and T. subterraneum (Subterranean Clover) were all found. Growing with them was a small patch of Chamaemelum nobile (Chamomile), now confined to three sites in Dorset. A little further on was Eleocharis quinqueflora (Few-flowered Spike-rush), but it was too early for Radiola linoides (Allseed) and Anagallis minima (Chaffweed). On the heath itself there was little flowering, except some bushes of Genista anglica (Petty Whin), but the verges had Geum rivale (Water Avens), Carex disticha (Brown Sedge) and Genista tinctoria (Dyer’s Greenweed). Some energetic members ventured into a bog to look at vegetative Deschampsia setacea (Bog Hair-grass), but again the heathland specialities such as Gentiana pneumonanthe (Marsh Gentian) and Rhynchospora fusca (Brown Beak-sedge) were not yet visible. A flush had Carex pulicaris (Flea Sedge), Cirsium dissectum X palustre, Dactylorhiza praetermissa (Southern Marsh-orchid), Ophioglossum vulgatum (Adder’s-tongue) and Sanguisorba officinalis (Great Burnet). The return to Dorchester was by way of Oakers Wood where there was still a good show of Pulmonaria longifolia (Narrow-leaved Lungwort) along the roadside verge. Another smaller party of members met at West Bexington, at the westward end of Chesil Beach. Crambe maritima (Sea Kale) was in full flower, along with some Glaucium flavum (Yellow-horned Poppy). The party was then invited to tea by Mr and Dr Eden, who have an extremely interesting Mediterranean shingle garden (with a good reptile fauna too). Many scarce plants such as Lathyrus aphaca (Yellow Vetchling), Vicia parviflora (Slender Tare) and Petroselinum segetum (Corn Parsley) are garden weeds here, and are well looked after. SUNDAY 18 MAY, PORTLAND AND FERRYBRIDGE A very large group, augmented by a few members of the Dorset Natural History and Archaeological Society, who had been our hosts the day before, met at Portland Heights. With the overnight mist 142 REPORT clearing one party went to St George’s Reforne, where there is a large churchyard with Arum italicum subsp. neglectum (Italian Lords-and-Ladies); many naturalised plants including Narcissus tazzetta (Bunch-flowered Daffodil), Scilla peruviana (Portuguese Squill) and Gladiolus communis (Eastern Gladiolus) were seen. Across the road is one of the quarry faces that holds Adiantum capillus-veneris (Maidenhair Fern). This party then proceeded to Church Ope Cove, a well-known botanical site. Here Valerianella eriocarpa (Hairy-fruited Cornsalad), Polypodium cambricum (Southern Polypody), Linum bienne (Pale Flax) and Orobanche hederae (Ivy Broomrape) were seen along with some small patches of Limonium recurvum (Portland Sea-lavender). No Gentianella anglica (Early Gentian) plants were found, but on the way back a native site for Adiantum capillus-veneris (Maidenhair Fern) was shown. The other party drove to Portland Bill. Parapholis incurva (Curved Hard-grass) grows on bare ground by the lighthouse and nearby is a long established clump of Borago officinalis (Borage). Walking eastwards along the coast there was much Crithmum maritimum (Rock Samphire), Inula crithmoides (Golden Samphire) and a magnificent stand of Salvia verbenaca (Wild Clary). The locus classicus of Limonium recurvum (Portland Sea-lavender) was admired, and then the party turned into one of the arable strips, where Medicago polymorpha (Toothed Medick) was found and Trifolium squamosum (Sea Clover). In the next field there was a mass of Lathyrus aphaca (Yellow Vetchling) and Ranunculus parviflorus (Small-flowered Buttercup). A walk behind the Pulpit Inn to find Gentianella anglica (Early Gentian) was unsuccessful, but here there was much Thesium humifusum (Bastard Toadflax) and a rich calcareous flora. Lunch was held back at Portland Heights, where Valerianella eriocarpa (Hairy-fruited Cornsalad) was seen again. Then the group descended to Ferrybridge, where Chesil Beach joins the Isle of Portland. Despite roads, railways, car parks, visitors and windsurfers this is still an extremely rich site and the mass of Armeria maritima (Thrift) in flower made it extremely colourful. Here the party saw Polycarpon tetraphyllum (Four-leaved Allseed), first recorded here in 1770, with Phleum arenarium (Sand Cat’s-tail), Trifolium scabrum (Rough Clover) and Vulpia fasciculata (Dune Fescue). The car park held Anthriscus caucalis (Bur Chervil). Across the road, by the old railway line, was what looked like Limonium recurvum with L. dodartiforme, Calystegia soldanella (Sea Bindweed) and Eryngium maritimum (Sea Holly). Further on was a newly discovered very large patch of Asparagus officinalis subsp. prostratus (Wild Asparagus), which, like the Polycarpon tetraphyllum, has been known in the area for over 200 years. Euphorbia portlandica (Portland Spurge) and Suaeda vera (Shrubby Sea-blite) were also frequent here, along with a few plants of Papaver dubium subsp. lecoqgii (Long-headed Poppy). Further on there were found huge drifts of Anisantha madritensis (Compact Brome), looking very fresh and reddish-green, and finally many plants of Geranium purpureum (Little Robin) and G. rotundifolium (Round-leaved Crane’s-bill). Ice creams and tea rewarded those not already surfeited! Davip PEARMAN B.S.B.I. PUBLICATIONS HANDBOOKS Each handbook deals in depth with one or more difficult groups of British and Irish plants. No. 1 SEDGES OF THE BRITISH ISLES A.C. Jermy, A. O. Chater and R. W. David. 1982. 268 pages, with a line drawing and distribution map for every species. Paperback. ISBN 0 901158 05 4. No. 2 UMBELLIFERS OF THE BRITISH ISLES T. G. Tutin. 1980. 197 pages, with a line drawing for each species. Paperback. ISBN 0 901158 02 X. Reprinting. No. 3 DOCKS AND KNOTWEEDS OF THE BRITISH ISLES J. E. Lousley and D. H. Kent. 1981. 205 pages, with many line drawings of native and alien taxa. Paperback. ISBN 0 901158 04 6. Out of print: new edition in preparation. No. 4 WILLOWS AND POPLARS OF GREAT BRITAIN AND IRELAND R. D. Meikle. 1984. 198 pages, with 63 line drawings of all species, subspecies, varieties and hybrids. Paperback. ISBN 0 901158 07 0. No.5 CHAROPHYTES OF GREAT BRITAIN AND IRELAND J. A. Moore. 1986. 144 pages, with line drawings of 39 species and 17 distribution maps. Paperback. ISBN 0 901158 16 X. No. 6 CRUCIFERS OF GREAT BRITAIN AND IRELAND T.C. G. Rich. 1991. 336 pages, with descriptions of 140 taxa, most illustrated with line drawings and 60 with distribution maps. Paperback. ISBN 0 901158 20 8. No. 7 ROSES OF GREAT BRITAIN AND IRELAND G. G. Graham and A. L. Primavesi. 1993. 208 pages, with descriptions and illustrations of 12 native and eight introduced species, and descriptions of 83 hybrids. Distribution maps are included of 31 selective species and hybrids. Paperback. ISBN 0 901158 22 4. No. 8 PONDWEEDS OF GREAT BRITAIN AND IRELAND - C. D. Preston. 1995. 350 pages. Covers 21 species and 26 hybrids with many full page illustrations including habit and details of leaves, stems and stipules and up-to-date distription maps. Includes full descriptions, keys, and introduction with accounts of history, evolution and life histories. Paperback. ISBN 0 901158 24 0. No.9 DANDELIONS OF GREAT BRITAIN AND IRELAND A. A. Dudman & A. J. Richards. 1997. 344 pages. 235 species described with silhouettes of herbarium specimens. 179 distribution maps. Paperback. ISBN 0 901158 25 9. OTHER PUBLICATIONS LIST OF VASCULAR PLANTS OF THE BRITISH ISLES D. H. Kent, 1992. 400 pages. Nomenclature and sequence followed by Stace in New Flora of the British Isles. Paperback. ISBN 0 901158 21 6. ATLAS OF THE BRITISH FLORA F. H. Perring and S. M. Walters, 1990. Reprint of 3rd Edn, 1982. 468 pages. Distribution maps of over 1700 species including updated maps for 321 Red Data Book species. New bibliography of updated distribution maps published elsewhere since 1st Edn, 1962. Paperback. ISBN 0 901158 19 4. ALIEN PLANTS OF THE BRITISH ISLES E. J. Clement and M. C. Foster, 1994. 590 pages. Lists 3586 species found growing wild and includes frequency of occurrence, areas of origin, means of introduction, references to descriptions and illustrations and location of herbarium material. Excludes grasses. Paperback. ISBN 0 901158 23 2. ALIEN GRASSES OF THE BRITISH ISLES T. B. Ryves, E. J. Clement and M. C. Foster, 1996. 181 pages including 29 of line drawings by G. M. S. Easy. Lists 700 species. Format and contents as for Alien plants. Paperback. ISBN 0 901158 27 5. ENGLISH NAMES OF WILD FLOWERS J. G. Dony, S. L. Jury and F. H. Perring, 1986. 2nd Edn. 123 pages. A list recommended by the BSBI arranged alphabetically, Latin-English and English-Latin. Paperback. ISBN 0 901158 15 1. Available from the official agents for B.S.B.I. Publications: F. & M. Perring, Green Acre, Wood Lane, Oundle, Peterborough PE8 5TP, England. Tel: 01832 273388 Fax: 01832 274568 BOTANICAL SOCIETY OF THE BRITISH ISLES (B.S.B.1.) BoBl The B.S.B.I. was founded in 1836 and has a membership of 2,850. It is the major source of information on the status and distribution of British and Irish flowering plants and ferns. This information, which is gathered through a network of county recorders, is vital for the conservation of the plants and is the basis of the Red data books for vascular plants in Great Britain and Ireland. The Society arranges conferences and field meet- ings throughout the British Isles and, occasionally, abroad. It organises plant distribution surveys and publishes plant atlases and handbooks on difficult groups such as sedges and willows. It has a panel of referees available to members to name problem plants. Through its Conservation Committee it plays an active part in the protection of our threatened plants. It welcomes all botanists, professional and amateur alike, as members. Details of membership and any other information about the Society may be obtained from: The Hon. General Secretary, Botanical Society of the British Isles, c/o Department of Botany, The Natural History Museum, Cromwell Road, London, SW7 5BD. INSTRUCTIONS TO CONTRIBUTORS Scope. Authors are invited to submit Papers and Notes concerning British and Irish vascular plants, their taxonomy, biosystematics, ecology, distribution and conservation, as well as topics of a more general or historical nature. Authors should consult the Hon. Receiving Editor for advice on suitability or any other matter relating to submission of manuscripts. Papers and Notes must be submitted in duplicate, typewritten on one side of the paper, with wide margins and double-spaced throughout. Submission of copy on computer discs is acceptable but must also be accompanied by two hard copies of the text, double spaced throughout. Format should follow that used in recent issues of Watsonia. Underline where italics are required. Names of periodicals should be given in full, and herbaria abbreviated as in British and Irish herbaria (Kent & Allen 1984). The Latin names and English names of plants should follow the New Flora of the British Isles, 2nd ed. (Stace 1997). Further details on format can be obtained from the Hon. Receiving Editor. Tables, figure legends & appendices should be typed on separate sheets and attached at the end of the typescript. Figures should be drawn in black ink or be laser-printed and identified in pencil on the back with their number and the author’s name. They should be no more than three times final size, bearing in mind they will normally be reduced to occupy the full width of a page. Scale-bars are essential on plant illustrations and maps. Lettering should be of high-quality and may be done in pencil and left to the printer. Black and white photographs can be accepted if they assist in the understanding of the article. Contributors must sign a copyright declaration prior to publication which assigns the copyright of their material to the Botanical Society of the British Isles. Twenty-five offprints are given free to authors of Papers, Notes and Obituaries; further copies may be purchased in multiples of 25 at the current price. The Society takes no responsibility for the views expressed by authors of Papers, Notes, Book Reviews or Obituaries. Submission of manuscripts Papers and Notes: Dr B. S. Rushton, School of Applied Biological and Chemical Sciences, University of Ulster, Coleraine, Co. Londonderry, N. Ireland, BTS2 1SA. Books for Review: Dr C. D. Preston, Biological Records Centre; Monks Wood, Abbots Ripton, Huntingdon, PE17 2LS. Plant Records: the appropriate vice-county recorder, who should then send them to Dr C. D. Preston, Biological Records Centre, Monks Wood, Abbots Ripton, Huntingdon, PE17 2LS. Obituaries: Mrs M. Briggs, 9 Arun Prospect, Pulborough, West Sussex, RH20 1AL. Back issues of Watsonia are handled by Dawson UK Limited, Cannon House, Folkestone, Kent, - CT19 SEE to whom orders for all issues prior to Volume 21 part 1 should be sent. Recent issues (Vol. 21 part 1 onwards) are available from Mr M. Walpole, B.S.B.I., 68 Outwoods Road, Loughborough, Leicestershire, LE11 3LY. Watsonia February 1998 Volume twenty two Part one Contents Rumsey, F. J., JERMy, A. C. & SHEFFIELD, E. The independent gametophytic stage of Trichomanes speciosum Willd. (Hymenophyllaceae), the ee Fern and its distribution in the British Isles PEARMAN, D, A”, PRESTON, ©. D2; Roy, D2 Bec& SHEITART. A The. use of B. S-B: 1 Monitoring Scheme data to predict nationally scarce species in Britain WILKINSON, D. M. Sere between species richness and ie in Welsh aquatic floras DANIELS, R. E. MeDomEn. iE Ue & Rusu, K F. The current status © Rumex rupestris Le Gall (Polygonaceae) in England and Wales, and threats to its survival and genetic diversity .. Drist, Ro, ©. The distnbution anal status of Conpnephons canescens (L.) P. Beauv. (Poaceae) in Britain and the Channel Islands with particular reference to its conservation KITCHENER, G. D. & McKean, D. R. Hybrids of Epilobium brunnescens (Cockayne) Raven & me One ae and their occurrence in the - British Isles PRESTON: @~ De BAILEY. see & Ponamesaem, Pp. M. A reassessment nor the hybrid Potamogeton x gessnacensis G. Fisch. (P. natans X P. poe Potamogetonaceae) in Britain PRESTON, €. D., HOLLINGSwortH, P. Me & Goran R. i Potamonen pectinatus ie x P. vaginatus Turez. (P. X_ bottnicus ea a ane identified hybrid in the British Isles ALLEN, D. E. Five new species of Rubus L. (Rosaceae) wiih transmarine ranges Bult, A. L. Four new species of Rubus L. (Rosaceae) from eastern England ... TURNER ETTLINGER, D. M. A new variety of Ophyrs apifera Hudson (Orchidaceae) NOTES Corner, R. W. M. Northern limits attained by native British olanin in North Peary, Greenland ae Gulliver, R. L. Population sizes a Genhanella ulanteed (will ) Bocnice Dune Gentian, on Colonsay (v.c. 102) in 1996 page P. & Stace, C. A. X Agropogon robaisonn (nice) Meldent & Dee. McC@hlint: Rose, F. Gentianella uliginosa (Willd. ) Boomer (Gentanaeese) fourein in Colones (v. c. 102), new to Scotland Book REVIEWS OBITUARIES REPORT Published by the Botanical Society of the British Isles ISSN 0043-1532 Filmset by WILMASET LTD, BIRKENHEAD, MERSEYSIDE Printed in Great Britain by EATON PRESS LIMITED, WESTFIELD ROAD, WALLASEY.:MERSEYSIDE 1) 21-28 29-32 33-4) 41-47 49-60 61-68 69-82 83-96 97-104 105-107 109-110 Pia Ms 113-114 114-116 117-125 127-137 139-142 A Botanical Society of the British Isles , J-R. Edmondson, D.L. Kelly tere f On =| Fi r N Bp YU iS | Bes < O N NE E = Pa = Ae ie < 52 (eS) = C.D. Preston, B.S. Rushton, M.N. Sanford Editors Botanical Society of the British Isles Patron: Her Majesty Queen Elizabeth the Queen Mother Applications for membership should be addressed to the Hon. General Secretary, c/o Department of Botany, The Natural History Museum, Cromwell Road, London, SW7 5BD, from whom copies of the Society’s Prospectus may be obtained. Officers for 1998-99 President, Mrs M. Briggs Vice-Presidents, Mr R. G. Ellis, Dr C. D. Preston, Dr R. J. Gornall, Mr M. Walpole Honorary General Secretary, Mr R. G. Ellis Honorary Treasurer, Mr M. E. Braithwaite Editors of Watsonia Papers and Notes, J. R. Edmondson, D. L. Kelly, B. S. Rushton, M. N. Sanford* Plant Records, C. D. Preston Book Reviews, C. D. Preston Obituaries, M. Briggs *Receiving editor, to whom all MSS should be sent (see inside back cover). © 1998 Botanical Society of the British Isles The Society takes no responsibility for the views expressed by authors of Papers, Notes, Book Reviews or Obituaries. Accredited with the International Association for Plant Taxonomy for the purpose of registration of new non-fungal plant names. The cover illustration of Meconopsis cambrica (L. Viguier (Welsh Poppy) was drawn by Rosemary Wise. Watsonia 22: 143-152 (1998) 143 Chromosome numbers and flower sizes of 4 7” Roth. and Ulex gallii Planch. in Dorset J. M. BULLOCK, J. CONNOR, S. CARRINGTON and R. J. EDWARDS N.E.R.C. Institute of Terrestrial Ecology, Furzebrook Research Station, Wareham, Dorset, BH20 5AS ABSTRACT This study was carried out to address the continuing uncertainty concerning the identification and chromosome numbers of the closely-related species Ulex minor Roth. and U. gallii Planch. (Fabaceae). The species co-occur in Dorset and chromosome counts from mixed and single-species populations gave results of n = 16 for all 52 U. minor plants sampled and of n = 32 for 52 U. gallii plants. The U. minor count is uncontroversial. The U. gallii count has been reported in other studies, but some studies in Great Britain, Ireland, France and Spain have reported n = 48. One plant, found in a mixed population of the two species and identified as U. gallii, had a count of n = 24. This number has never been reported before for a European Ulex and the plant may be a U. gallii x U. minor hybrid. It had an intermediate flower size. Although the two species are best distinguished by flower size, a survey over Dorset showed appreciable overlap in flower standard and calyx lengths. However, use of a suite of characters (flower size, spine length and bush size) always gave correct identification when tested against chromosome number. A search for hybrids — based on the hypothesis that mixed populations will contain more intermediate plants (because of hybridization) than single- species populations — suggested that U. gallii x U. minor hybrids are rare. KEYWORDS: Ulex gallii x U. minor, U. gallii x U. europaeus, gorse, morphometrics, ploidy levels. INTRODUCTION Ulex minor Roth. and U. gallii Planch. (Fabaceae) have caused problems to British and European botanists since U. gallii was first described by Planchon in 1849. Before its description, U. gallii was taken to be a large form of U. minor, and there are still problems in distinguishing the two species (Castroviejo & Valdés-Bermejo 1990; Gloaguen 1986; Proctor 1967). This is because the species show substantial overlap in many characters, e.g. bush height, spine length, flower colour and pod size (Proctor 1965). Although flower size (standard and calyx lengths) is the most reliable character for species identification, even this character is not completely distinct between the species (Proctor 1965). Chromosome counts might be expected to provide a method for separating the species; indeed, they do have different chromosome numbers (see below). However, cytological studies over the past decade have opened up a new area of controversy and debate. Several studies of U. minor in France and Spain have all arrived at a count of 2n = 32 (e.g. Alvarez Martinez et al. 1988; Castroviejo & Valdés-Bermejo 1983, 1990; de Castro 1941; Fernandez Prieto et al. 1993; Misset 1990; Misset & Gourret 1996). The same results were obtained from a site in Dorset by Fernandez Prieto et al. (1993) and from a site in Surrey listed in the B.S.B.I. database. However, chromosome counts for U. gallii have been more varied. The first count, by de Castro (1943), suggested 2n = 80. This was probably wrong, and the debate over the last 15 years has revolved around the fact that studies of U. gallii in Europe have found both 2n = 96 and 2n = 64. Castroviejo & Valdés-Bermejo (1983, 1990) counted both 2n = 64 and 96 in Spain. Alvarez Martinez and co-workers (Alvarez Martinez et al. 1988; Fernandez Prieto et al. 1993) found 2n = 64 in Spain and France, 2n = 96 in Spain and one plant with 2n = 32 in Spain. Misset (1990; Misset & Gourret 1996) found 2n = 96 in north-west France, although she had one example of 2n = 64. In the British Isles, Fernandez Prieto ef al. (1993) studied eight - U. gallii sites in Devon and Cornwall and found only plants with 2n = 64. However the B.S.B.I. database contains counts of 2n = 96 from three U. gallii sites, on Alderney, in Derbyshire and in County Dublin. Where methods have been given, all these studies distinguished species in the field using the differences of flower calyx and standard lengths, spine length and bush size reported by Proctor (1965). Bf fe ir id Big s & +s xe Tite . LAs iY et Sw Ea pod, ¢ i 144 ‘A J. M. BULLOCK, ‘i ONO S. CARRINGTON AND R. J. EDWARDS This, confusion! nds duompled suggestions for changes in the taxonomy of the two species, involving splitting.U. gallii Planch according to chromosome numbers. Castroviejo & Valdés-Bermejo (1983, 1990) suggested that plants. with 2n*= 64 should be named U. minor subsp. breoganii and those with 2n = 96 become U. minor subsp. ‘gallii (with U. minor becoming U. minor subsp. minor). Conversely, Alvarez Martinez et al. (1988) suggested 2n = 64 plants should remain U. gallii while 2n = 96 plants become U. cantabricus. However, neither suggestion has been accepted generally. Clearly, more data are needed on the chromosome numbers of U. gallii and U. minor. Another unresolved question concerns the occurrence of U. gallii x U. minor hybrids. While U. gallii x U. europaeus hybrids are commonly described (Benoit 1962; Gloaguen 1986; Misset & Fontenelle 1992; Stace 1975) (interestingly, no U. minor x U. europaeus hybrids have been suggested), the evidence for U. gallii x U. minor hybrids is weak (see Stace 1975). Millener (1952) failed to produce FIGURE |. The distributions of Ulex minor and U. gallii in Britain. U. minor occurs east of the line ................ , and U. gallii occurs west of the line - - --—- and in East Anglia in the region marked + CHROMOSOME NUMBERS OF ULEX MINOR AND U. GALLII 145 TABLE 1. CHARACTERS OF TWO ULEX SPECIES IN BRITAIN, AS DESCRIBED BY PROCTOR (1965) Ulex minor Ulex gallii Flowering season: July-September Flowering season: as U. minor Flowers Flowers Standard: 6-12 mm (mode 9:5 mm) Standard: 10:-5—18 mm (mode 15 mm) Calyx: 6-9 mm Calyx: 9:5-12°5 mm Wings: variable Wings: strongly curved, c. | mm longer than keel Pedicels: 3-5 mm, appressed hairs Pedicels: as U. minor Flower colour — mostly aureolin and lemon yellow =‘ Flower colour — mostly buttercup and Indian yellow Bracteoles: 0:5—0:8 x 0:6-0°8 mm Bracteoles: as U. minor Pod: 6-11 mm Pod: 6-12 mm Spines: 6-25 mm Spines: 8-34 mm Bush height: 5-150 cm Bush height: 10-200 cm Other: flower opens more widely than U. minor after pollination. Possible red veining of standard. pods or seeds in an attempted cross between U. minor (3) and U. gallii (2). All suggestions of U. gallii x U. minor hybrids in the field derive from notes by Corillion (1950) and Lambinon (1962) on plants with vegetative and floral characteristics intermediate between the two species. Given the overlap in characteristics of the described species (e.g. Proctor 1965), it is extremely tenuous to call intermediate plants hybrids. Chromosome counts would provide more solid evidence of hybrids, if plants of intermediate chromosome numbers were found in mixed populations of U. gallii and U. minor. U. gallii occurs in the west of Britain and U. minor is in the south-east, and their distributions only overlap substantially in Dorset (Fig. 1). Within Dorset most heaths contain only one of the two Ulex species (single-species populations), although there are several heaths (mixed populations) with both species (unpublished data J. M. Bullock & R. J. Edwards); this suggests competitive separation of the two species. It is therefore likely that any British U. gallii x U. minor hybrids would be found in Dorset. In this study we carried out chromosome counts of a large sample of plants from a mixed population of U. gallii and U. minor and from single-species populations in Dorset in order to answer the following questions. 1. What are the chromosome numbers of U. gallii and U. minor in Dorset? 2. Is field identification of U. gallii and U. minor using Proctor’s (1965) criteria reliable, as tested by chromosome counts? 3. Is there any chromosomal evidence for U. gallii x U. minor hybrids in mixed populations? Questions 2 and 3 were explored further by taking measurements of floral standard and calyx lengths from U. gallii and U. minor plants in mixed and single-species populations throughout Dorset in order: 1. to assess the overlap in flower size (an extension of Proctor’s 1965 work); and 2. to determine whether mixed populations have a greater proportion of plants with flower sizes intermediate between the two species; this would provide indirect evidence of hybrids, which (if they exist) should be more prevalent in mixed populations. Proctor (1965) showed that standard and calyx lengths were the best characters with which to separate the species (Table 1). Therefore, if hybrids have intermediate characters, then standard and calyx lengths should be the best characters with which to identify hybrids. It should be noted that this hypothesis does not assume that all intermediate plants are hybrids, but does assume that the presence of hybrids will change the frequency distribution of flower sizes. The third distinguishing character by Proctor (1965) was flower colour, but we did not use this because it is a discontinuous variable and would be difficult to use to identify intermediates. METHODS CHROMOSOME COUNTS A survey of the occurrence of both species in each heath in Dorset had been carried out in 1995 (R. J. Edwards, unpublished data), and this was used to select three heaths from which to sample. Gore Heath 146 J. M. BULLOCK, J. CONNOR, S. CARRINGTON AND R. J. EDWARDS TABLE 2. THE DORSET HEATHS (WITH GRID REFERENCE) FROM WHICH SAMPLES OF ULEX SPP. FLOWER SIZES WERE TAKEN U. minor only heaths Arne Heath SY/964.878 Bovington Heath SY/838.916 Hurn Common SZ/136.960 Winfrith Heath SY/380.876 U. gallii only heaths Canford Heath SZ/02 1.962 Cripplestyle Common SU/092.118 Puddletown Forest SY/728.920 Upton Heath SY/974.956 Mixed heaths Ferndown Common SZ/064.994 Godlingston Heath SZ/015.820 Gore Heath SY/924.900 Ham Common SY/976.908 Holt Heath SU/054.030 Parley Common SZ/084.986 Stoborough Heath SY/936.848 Studland Heath SZ/022.844 (S Y/924.900) contained both species and the populations were intermingled. This was judged to be the best example of a mixed population in Dorset. Plants were sampled at random over the whole heath. Each plant was identified in the field as U. minor or U. gallii using Proctor’s (1965) characters; the most useful characters were flower size (i.e. standard length), spine length and bush size. These characters were generally larger in U. gallii than in U. minor and we used field measurement of all three characters to distinguish the species (we did not make notes of these measurements). Every plant encountered was sampled, even those which were difficult to identify. Buds were taken from 50 U. minor bushes and the same number of U. gallii bushes. To provide baseline chromosome counts for both species, Ulex plants were sampled in heaths containing single-species populations. Ten U. gallii bushes were sampled from Canford Heath (SZ/021.962) and ten U. minor bushes were sampled from Winfrith Heath (SY/ 380.876). All sampling was done on 12-13 August 1996. The methods used for counting chromosomes were the same as those given by Misset (1990) and Fernandez Prieto et al. (1993) (see also Gurr 1965). Flower buds .of c. 2 mm length were fixed in the field in Carnoy’s fixative (3:1 Glacial acetic acid-Ethanol) and then kept refrigerated for at least 48 hours. The anthers were then dissected out on a microscope slide in a drop of aceto-carmine and squashed under a coverslip. Counts were made of stained chromosomes in pollen cells at metaphase 1. Counts were made from at least two buds from each plant. FLOWER SIZES The 1995 survey was used to select four heaths which contained only U. minor, another four heaths which contained only U. gallii, and eight heaths with mixed populations (Table 2). In each heath 20 bushes of each of the gorse species present (identified using Proctor’s characters) were chosen at random over the entire extent of the heath. As with the chromosome counts, difficult bushes were not avoided. Five fully opened flowers were picked from each bush and stored in an ice-box. Within 24 hours of collection the floral standard and calyx lengths of each flower was measured to the nearest 0:5 mm. The five measures for each bush were used to calculate mean standard and calyx lengths for each of the 320 bushes sampled. Sampling was carried out over August 1995. RESULTS CHROMOSOME COUNTS The ten U. minor plants from Winfrith Heath all gave chromosome counts of n= 16 and the ten U. gallii plants from Canford Heath had counts of n = 32. Of the plants from the mixed population on Gore Heath, CHROMOSOME NUMBERS OF ULEX MINOR AND U. GALLII 147 all those identified in the field as U. minor gave counts of n = 16 (counts were taken from 42 plants; we were unable to obtain adequate preparations from the remaining eight plants), but of those identified as U. gallii 42 gave counts of n = 32 and one gave a count of n = 24 (Fig. 2). To assess the consistency of this last unusual chromosome number, counts were made from ten buds on the plant and all gave n = 24. Several preparations were made of pollen grains, and the pollen for the n = 24 plant appeared normal and similar to that of the n= 16 and n = 32 plants. A single flower preserved with the buds of the n = 24 plant had a standard and a calyx length of 11-5 mm and 8-5 mm respectively. This plant had been identified in the field as U. gallii on the basis of its standard and spine lengths and bush size. However, we were not able to relocate this plant in the field in order to make detailed measurements of its floral and vegetative characters. FLOWER SIZES T-tests showed that the average floral standard and calyx lengths of the U. minor and U. gallii were significantly different (single-species populations, standard length, t = 19-4, p<0-001, calyx length, t = 16-3, p<0-001; mixed populations, standard length, t = 28-6, p<0-001, calyx length, t = 25:8, p<0-001), and most of the U. gallii plants had longer standards and calyces than any U. minor plant (Figs 3 & 4). However there was an overlap in flower sizes, with both species having bushes with mean standard lengths of 11—-12-5 mm and mean calyx lengths of 8-10 mm. The distribution of flower sizes of each species did not differ between the single- species populations and the mixed populations (standard length, U. minor ? = 12: 3s df = 7, nsd; U. gallii y° = 6-6, df = 8, nsd; calyx length, U. minor x? = 10-7, df = 6, nsd; U. gallii 7 = adie—s6. nse): The hypothesis that mixed populations would have a greater proportion of bushes with intermediate flower sizes (suggesting the presence of U. gallii x U. minor hybrids) was tested by comparing the proportions of three categories of bushes between the single-species and mixed populations. These categories were: bushes with intermediate flower sizes, bushes with smaller than intermediate flower sizes and bushes with larger than intermediate flower sizes. Two definitions of intermediate standard size were tested: a. 11: a= 12 mm, b. 11—12:5 mm. Definition a gave no significant differences between the population types (y* = 4-3, df = 2, nsd). Definition b resulted in a significant difference (y = 6°5, df = 2, p<0-05), but this was because there was a smaller rather than greater proportion of intermediate plants in the mixed (16%) than in the single-species population (25%). Similarly, two definitions of intermediate calyx size were tested: a. 8:5—9: > mm, b. 8-10 mm. Neither definition gave a significant difference between the population types (a, v7 = 3-1, df = 2, nsd; b, y° = 3-8, df = 2, nsd). DISCUSSION CHROMOSOME NUMBERS OF U. GALLII AND U. MINOR Despite the overlap in flower sizes of the two Ulex species (Figs 3 & 4), the suite of character differences (we found the most useful to be flower size, bush height and spine length) between the two species allow accurate field identification of the species. There was complete agreement between the field identification of the plants and the chromosome counts obtained, n = 32 for U. gallii and n = 16 for U. minor (the single unusual count of n = 24 is discussed below), even for plants in an extensively mixed population. While there is no controversy about the chromosome numbers of U. minor (2n = 32), this paper contributes to the debate in France and Spain on the chromosome numbers of U. gallii (2n = 64 in this paper). Misset (1990; Misset & Gourret 1996) contended that U. gallii has 2n = 96 and called her single example of U. gallii with 2n = 64 a “ploidy accident’ (Misset & Gourret 1996). Castroviejo & Valdés- Bermejo (1990) also asserted that 2n = 96 for U. gallii and suggested that the counts by Alvarez Martinez et al. (1988) of 2n = 64 were mistakes. Alvarez Martinez and co-workers (Alvarez Martinez et al. 1988; Fernandez Prieto et al. 1993) however suggested that the two ploidy levels for U. gallii are’ common (even suggesting that U. gallii with 2n = 32 may occur). These workers have all identified their specimens using standard and calyx lengths (and sometimes other characters such as spine lengths), although most are not clear about the criteria used to distinguish the species. We assume that they have used the size range of flower sizes for each species reported by J. M. BULLOCK, J. CONNOR, S. CARRINGTON AND R. J. EDWARDS 148 ew? Ae 0g ] sseydejour ye s][aD ‘7 AUNDIA uino¢e CHROMOSOME NUMBERS OF ULEX MINOR AND U. GALLII 149 a) Standard lengths on single-species heaths 30 25 = 20 : = : & 4 ene ire E] Ulex minor 3 5 fie ee el Se QO) Ulex gallii ao “ “4 ao 5) eS (| (alse 2 5S a Oe S10 5 0 b) Standard lengths on mixed heaths 30 25 20 % frequency Flower standard length (mm) FIGURE 3. Histograms of the frequency distributions of flower standard length (using plant means) for the two Ulex species in single species populations (80 plants sampled for each species) and in mixed populations (160 plants sampled for each species). Proctor (1965) (U. minor standard = 6-12:5 mm, calyx = 5-5—10-°5 mm; U. gallii standard = 10-:5-18 mm, calyx = 8-5—14:5 mm), which are similar to those found in this study (U. minor standard = 7:5- 12:5 mm, calyx = 6-10 mm; U. gallii standard = 11-17 mm, calyx = 8-13-5 mm). However, where measurements have been reported, they do not help to clarify the taxonomic problems concerning U. gallii ploidy levels. Alvarez Martinez et al. (1988) reported standard and calyx lengths of 8-7—-15-8 mm and 6-8-14:5 mm respectively for plants with 2n = 64 identified as U. gallii and of 10-3—-14:3 mm and 8-7—12-4 mm for plants with 2n = 96 named as U. cantabricus. These are curious results because the minimum lengths for U. gallii are very low in comparison to those in this 150 J. M. BULLOCK, J. CONNOR, S. CARRINGTON AND R. J. EDWARDS a) Calyx lengths on single-species heaths % frequency Seat feo. £ Ome 8:9) ay 9-5 e108 AOS alt ol :5) 128 125 aS 13.5 b) Calyx lengths on mixed heaths % frequency Flower calyx length (mm) FIGURE 4. Histograms of the frequency distributions of flower calyx length (using plant means) for the two Ulex species in single species populations (80 plants sampled for each species) and in mixed populations (160 plants sampled for each species). and Proctor’s studies, and the U. cantabricus measurements fall within the range of the putative U. gallii plants in these other studies. Castroviejo & Valdés-Bermejo (1990) gave calyx lengths of 7—9-5 mm for plants identified as U. minor with 2n = 32, 9-514 mm for plants identified as U. gallii with 2n = 96, and 8-5—11 mm for plants named as U. minor subsp. breoganii with 2n = 64. Therefore, the plants named as U. minor subsp. breoganii have calyx lengths intermediate between those given for U. minor and U. gallii and Castroviejo & Valdés-Bermejo (1990) describe this putative subspecies as having a generally intermediate morphology. It seems likely that U. gallii has two ploidy levels on the European mainland and — given the three CHROMOSOME NUMBERS OF ULEX MINOR AND U. GALLII 151 examples with 2n = 96 on the B.S.B.I. database and the populations with 2n = 64 found by Fernandez Prieto et al. (1993) and in this study — possibly in the British Isles. However, taxonomic and morphological confusion means that it is unclear whether it is possible to distinguish these ploidy levels morphologically or what their taxonomic standing should be. Clearly, there is a need to address these issues by extensive morphological and cytological studies in Britain and Europe. It would seem premature to adopt the names U. cantabricus or U. minor subsp. breoganii for one or other of the ploidy levels. U. GALLII X U. MINOR HYBRIDS The mechanism that produced the single Ulex bush with a count of n = 24 (2n = 48) can only be guessed at. One possible explanation is that it was a U. gallii x U. minor hybrid, and indeed it did have an exactly intermediate chromosome number. Other supporting evidence (although tenuous) is that the single flower retrieved from the specimen was of intermediate size. It is usually stated or implied that the base chromosome number (x) for the genus Ulex is 16, and that U. minor is diploid, U. gallii is tetraploid (2n = 64) or hexaploid (2n = 96) and that U. europaeus is hexaploid (Castroviejo & Valdés- Bermejo 1990; Fernandez Prieto et al. 1993; Misset & Gourret 1996). If x = 16 then a plant (whether hybrid or not) with 2n = 48 would be a triploid, and would usually be infertile (Felber & Bever 1997). However, microscopic examination of pollen grains from this plant suggested that they were fertile. If the base number of Ulex was x = 8 (as suggested by de Castro 1941, 1943) then the putative hybrid would be hexaploid with tetraploid (U. minor) and octoploid (U. gallii) parents. If the n = 24 plant was a U. gallii x U. minor hybrid, then such hybrids are extremely uncommon. This chromosome count has never before been reported for U. minor, U. gallii or U. europaeus, or any other European Ulex. Only one such plant was found in a sample of 85 bushes from Gore Heath; a heath where the intimate mixing of the two species should provide ideal conditions for hybridisation. The fact that mixed heaths did not have a greater proportion of bushes with intermediate flower sizes (standard and calyx lengths) than single-species heaths also provides indirect evidence against there being any great abundance of hybrids. While not all intermediate plants are necessarily hybrids (given the morphological overlap between the species), the presence of hybrids should increase the proportion of plants with intermediate characters. However, such a conclusion assumes that hybrids will have intermediate characters. This assumption has been shown to be only partly true for plant hybrids; certain characters may be indistinguishable from those of one parent or the other (Stace 1975; Riesberg & Ellstrand 1993). However, without better information, this assumption is acceptable and these data will be of use until there are more detailed studies which link morphological characters with allozyme or DNA markers in populations over a wide geographical area. The possibility that there are few U. gallii x U. minor hybrids may be useful for British botanists, as it suggests that separation of the two species will not be complicated by the presence of hybrids. However, it raises the question of why such similar species with identical flowering seasons do not hybridise. Investigations of cross-compatibility and pollinator behaviour may suggest solutions. U, GALLII X U, EUROPAEUS HYBRIDS Although they were not looked for explicitly, our study of summer-flowering Ulex plants did not provide any evidence of U. gallii x U. europaeus hybrids. Such hybrids are commonly described, but evidence is usually based on intermediate vegetative and floral characters (Gloaguen 1986; Millener 1952; Stace 1975). Only Misset & Fontenelle (1992) give reliable evidence which is based on differing isoenzyme systems of U. gallii and U. europaeus, and shows that putative hybrids have elements of both isoenzyme systems. U. europaeus is usually described as 2n = 96 (de Castro 1941, 1943; Castroviejo & Valdés-Bermejo 1983; Misset 1990), and counts from four bushes sampled on Gore Heath on 14 January 1997 gave the same number (V. Herrera, unpublished data). The only chromosome counts for U. gallii x U. europaeus hybrids have given 2n = 96, and these are in papers which give the count for U. gallii of 2n = 96 (Misset 1990; Misset & Gourret 1996). Where the count for U. gallii is 2n = 64, the count for hybrids should be 2n = 80 (interestingly, the count originally given. by de Castro (1943) for U. gallii). The flowering season of U. gallii x U. europaeus hybrids is described as extending over the seasons for both the parent species (Gloaguen 1986; Millener 1952; Stace 1975), so samples spread over a longer period than covered in this study may provide more solid evidence of hybridization. 152 J. M. BULLOCK, J. CONNOR, S. CARRINGTON AND R. J. EDWARDS ACKNOWLEDGMENTS Alan Raybould gave valuable advice on chromosome staining and Ralph Clarke advised on statistics. Alan Gray, Roger Daniels and an anonymous referee kindly commented on earlier drafts. REFERENCES ALVAREZ MARTINEZ, M. J., FERNANDEZ CaAsAbo, M. A., FERNANDEZ PRIETO, H., NAVA FERNANDEZ, S. & VERA DE LA PUENTE, M. L. (1988). El género Ulex en la Comisa Cantabrica. I. Ulex gr. gallii-minor. Candollea 43: 483-497. BENOIT, P. M. (1962). Ulex europaeus x gallii. Proceedings of the Botanical Society of the British Isles 4: 414-415. CASTROVIEJO, S. & VALDES-BERMEJO, E. (1983). Notas sobre los tojos Gallegos. Anales Jardin Botdnico Madrid 40: 73-81. CASTROVIEJO, S. F. L. S. & VALDES-BERMEJO, E. (1990). On the identity of Ulex gallii Planchon (Leguminosae). Botanical journal of the Linnean Society 104: 303-308. CORILLION, R. (1950). Phanérogames intéressantes pour la Bretagne. Bulletin de la Société Scientifique de Bretagne 25: 131-140. DE CASTRO, D. (1941). Algumas contagens de cromosomas no género Ulex L. (sensu lato). Agronomia lusitanica 3: 103-113. DE CASTRO, D. (1943). Contribuigao para o conhecimento carioldgica dos géneros Ulex L. Stauracanthus Link e Nepa Webb. Agronomia lusitanica 5: 243-249. FELBER, F. & BEVER, J. D. (1997). Effect of triploid fitness on the coexistence of diploids and tetraploids. Biological journal of the Linnean Society 60: 95-106. FERNANDEZ PRIETO, J. A., NAVA FERNANDEZ, S., VERA DE LA PUENTE, M. L., ALVAREZ MARTINEZ, M. J., D1AzZ, T. E., CASADO, M. A. F., FERNANDEZ-CARVAJAL, M. C. & VILLARIAS, M. I. G. (1993). Chromosome numbers and geographical distribution of Ulex gallii and U. minor (Leguminosae). Botanical journal of the Linnean Society 112: 43-49.. GLOAGUEN, J. C. (1986). Les ajoncs de Bretagne. Bulletin de la Societe Botanique de France — lettres botanique 133: 363-385. Gurr, E. (1965). The rational use of dyes in biology. Leonard Hill, London. LAMBINON, J. (1962). Note sur les Ulex du Massif Armoricain. Lejeunia 9: 64-66. MILLENER, L. H. (1952). Experimental studies on the growth forms of the British species of Ulex L. Ph.D. thesis, Pembroke College, Cambridge and the University of New Zealand. MissetT, M. T. (1990). Données caryologiques chez le genre Ulex L. (Papilionidae) dans le Massif Armoricain. Taxon 39: 630-635. Misset, M.T. & FONTENELLE, C. (1992). Protein relationships between natural populations of Ulex europaeus and U. gallii (Faboideae, Genisteae) and their hybrids. Plant systematics and evolution 179: 19-25. MisseT, M. T. & GourRreET, J. P. (1996). Flow cytometric analysis of the different ploidy levels observed in the genus Ulex L. Faboideae-Genisteae in Brittany (France). Botanica acta 199: 72-79. Proctor, M. C. F. (1965). The distinguishing characteristics and geographical distributions of Ulex minor and Ulex gallii. Watsonia 6: 177-187. Proctor, M.C. F. (1967). The British species of Ulex. Proceedings of the Botanical Society of the British Isles 6: 379-380. STACE, C. A. ed. (1975). Hybridization and the flora of the British Isles. Academic Press, London. (Accepted September, 1997) Watsonia 22: 153-161 (1998) 153 Morphometric variation in Irish Sorbus L. (Rosaceae) J. PARNELL Herbarium, School of Botany, Trinity College, Dublin 2, Ireland and M. NEEDHAM _Lorrha, Nenagh, County Tipperary, Ireland ABSTRACT Sorbus aria (L.) Crantz and S. hibernica E. F. Warb. are shown to differ from one another in a number of characteristics not previously noted, in particular, S. aria has a longer petiole, longer leaf blade with a more sharply pointed apex, greater number of leaf teeth and more widely spreading veins than S. hibernica. It does not appear that S. aria is more variable than S. hibernica in Ireland as is suggested generally by the literature. In addition, despite the very limited number of Sorbus devoniensis specimens available for study in Ireland, this species appears readily distinguishable from other, vegetatively similar, species of Sorbus on the basis of its longer petioles, higher number of leaf-teeth, more acutely pointed leaves and widely spaced veins. KEYWORDS: Sorbus aria, Sorbus hibernica, Sorbus devoniensis, multivariate analysis. INTRODUCTION Sorbus is a critical genus represented in Ireland by seven species (Webb, Parnell & Doogue 1996). Two species, Sorbus aucuparia L. (Rowan) and Sorbus intermedia (Pers.) Ehrh., are readily separable from the rest purely on vegetative characteristics (their leaves are, respectively, pinnate or deeply lobed rather than more or less entire or shallowly lobed); these two species will not be considered in detail further. The other five species — Sorbus anglica Hedl., S. aria (L.) Crantz, S. devoniensis E. F. Warb, S. hibernica E. F. Warb. and Sorbus rupicola (Syme) Hedl. are usually each placed in one of three aggregate species groups — S. anglica in S. intermedia agg., S. aria, S. hibernica and S. rupicola in S. aria agg. and S. devoniensis in S. latifolia agg. (Stace 1997). Like S. aucuparia, S. aria is a diploid; all other species are polyploids and probably apomictic (Proctor, Proctor & Groenhof 1989). Species in the S. intermedia agg. and S. latifolia agg. probably originated as hybrids of S$. aria and S. aucuparia and the non-Irish S. torminalis (L.) Crantz (Clapham, Tutin & Moore 1987). Undoubtedly the most problematic distinction in this group of species in Ireland is between S. aria and S. hibernica. When writing the key to Sorbus for Webb, Parnell & Doogue (1996) Parnell indicated that the most obvious vegetative distinction between these two species lies in the upswept leaf-teeth of S. aria (where the outer margin of each tooth is longer than the inner) whereas the leaf-teeth of S. hibernica are straight and symmetrical. Additionally it is clear that the density of the white indumentum on the undersurface of the leaves of S. hibernica is usually greater than in S. aria. Stace (1997) indicates that while the leaves of S. hibernica may have 10 or fewer pairs of lateral veins, S. aria always has at least 10 pairs of veins. However he rightly points out both leaves and fruits are required for identification by beginners. These criteria are often difficult to meet; very many specimens in herbaria or in the field and also many brought in for identification are sterile or lack fruit. Whilst the key in Webb, Parnell & Doogue (1996), which relies on vegetative characters only, does allow most Irish material to be keyed out accurately and consistently it is clear that it could be improved; however, any improvement can only come about through a systematic description of the variation in leaf form in S. hibernica which currently does not exist. The aforementioned difficulties are compounded by the 154 J. PARNELL AND M. NEEDHAM view of most authors (e.g. Clapham, Tutin & Moore 1987) that S. aria is a relatively variable species in comparison to its polyploid relatives. The present data-set was collected with the object of attempting to ascertain the differences, if any, in quantifiable leaf characters between the five species of Sorbus listed above, with special reference to the distinctions between S. aria and S. hibernica and to discover if there are any new distinctions between these species. MATERIALS AND METHODS In order to make a comprehensive survey, all material (146 sheets) of the genus Sorbus from the two largest Irish herbaria (DBN & TCD) was examined. In addition two extensive collections of new material were made from large populations of S. hibernica (20 trees from Coolbawn, County North Tipperary, Grid. ref. R/831.923; v.c. H10 and 44 trees from Kilbeggan, County Westmeath, Grid. ref. N/366.371; v.c. H23). The material of all taxa was from wild populations. The number of populations sampled in total was 148 (each herbarium sheet represented materials from a single plant and, as far as it was possible to determine, population). As a minimum, rarely two, and more usually six mature leaves were measured from each plant. Consequently the 148 populations were represented by 672 sets of leaf character measurements. The 672 measurements were then averaged so as to provide a mean value for each of the 148 populations. Of these only four plants assigned to S. anglica, another four of S. devoniensis and seven of S. rupicola were located. Characters (Table 1) were measured on each leaf (measurements were either in degrees or in mm or cm as appropriate). Unfortunately, as can be seen from Table 1 it proved impossible to effectively measure or code for the degree of upsweptness of the leaf teeth or the density of the indumentum. However it was essential to have some a priori means of assigning names to specimens and these two characters were used as the primary method whereby S. aria and S. hibernica were initially identified. The other taxa were assigned names on the basis of the key in Webb, Parnell & Doogue (1996). A number of authors refer to pairs of veins in the leaf (vide Stace 1997); however our experience suggests that veins are not always strictly TABLE 1. CHARACTERS OF SORBUS TAXA CODED FOR ANALYSIS IN PCA Character number and its abbreviation 1. The angle made between the left-hand side of the base of the leaf blade and the petiole (Angbotlf) 2. The angle made between the right-hand side of the base of the leaf blade and the petiole (Angbotrh) 3. The angle made between the third lateral vein from the base of the leaf blade and the midrib (Angof3ve) 4. The angle made between the fourth lateral vein from the base of the leaf blade and the midrib (Angof4ve) 5. The angle made between the margin of the left-hand side of the top of the leaf blade and the midrib (Angtoplf) 6. The angle made between the margin of the right-hand side of the top of the leaf blade and the midnb (Angtoprh) 7. The distance between the excursion points of the third and fourth lateral veins at the midrib (Distbet34) 8. The distance from the third lateral vein from the base of the leaf blade to the base of the leaf blade (Distbot3v) 9. The distance from the third lateral vein from the base of the leaf blade to the top of the leaf blade (Distfrm3v) 0. The distance from the insertion point of first tooth on left-hand side of base of the leaf blade to the base of the leaf blade (Distteel) 11. The distance from the insertion point of the first tooth on left-hand side of the base of the leaf blade to the base of the leaf blade (Distteer) 12. The length of the leaf blade, measured along its midrib (Leaflen) 13. The breadth of the leaf blade, measured at its point of maximal width (Leafwid) 14. The length of the leaf blade, measured along its midrib, to the widest point of the leaf blade (Leaflentwp) 15. The number of secondary veins on the leaf blade (Noveins) 16. The number of teeth present in the top centimetre of the leaf blade (Noteeto) 17. The number of teeth on the left-hand side of the leaf blade (Notthlhs) 18. The number of teeth on the right-hand side of the leaf blade (Notthrhs) 19. The length of the petiole (Petlen) 20. The width of the petiole (Petwid) MORPHOLOGICAL VARIATION IN IRISH SORBUS L. 155 paired and therefore we counted individual veins rather than pairs. With the sole exception of character number 20, the width of the petiole (Petwid), all characters were more or less normally distributed (17 values of >90% after regression of their normal probability plot values and with values for kurtosis and skewness usually<1). In a few cases (except Petwid) where the latter values were>1, r remained >90% and transformation to attain normality was therefore not attempted. Petwid had large values for kurtosis and skewness (9.5 and 2.8 respectively) and a r~ for regression of its normal probability plot values of 50%. A very large number of transformations were attempted for Petwid in an attempt to normalise its distribution; however no significant increase in normality could be obtained. Trial and error showed that exclusion of Petwid from the analyses undertaken had a minimal effect and therefore it was included in an untransformed state. A number of different types of analysis were undertaken on these data, and two fundamentally different techniques were used. Firstly the data were ordinated. The ordination technique chosen was Principal Components Analysis (PCA) as produced by Datadesk 5.0.1 (cf. Data Description Inc., Ithaca, New York). With all biological data sets PCA will extract as many summary axes as there are original variables in the data. However, only the first few of these axes represent effective summaries of the data and are non-trivial. The key question is — which axes are these? Unfortunately, this vital question is virtually ignored in the taxonomic literature. Many, if not most, workers use the heuristic Kaiser-Guttmann criterion to determine which PCA axes are of significance (Jackson 1993). Simply put, this translates into consideration being given only to those axes whose eigenvalues are greater than unity (1), all other lesser scoring axes being ignored. Jackson (1993) argues that this criterion is too lax and permits consideration of components which are trivial explicators of the total variation pattern and which should really be ignored. However, others disagree and some standard ecological texts suggest that where the pattern of the data is weak, where the data-set is exceptionally large or where the investigator is concerned with the preservation of inter-object distances (i.e. pattern) it is possible to obtain perfectly valid plots associated with weak eigenvalues (Legendre & Legendre 1983). Obviously there is no universally accepted solution to this problem. However some, which are non-arbitrary, are discussed by Jackson (1993). According to him the best available single test for the importance of specific eigenvalues in PCA is that of Frontier (1976) who showed that the decrease in the eigenvalues of sequentially extracted axes in PCA generally follows a “broken-stick’’ distribution-type, where a fixed length is broken at random into a number of segments. Jackson (1993) uses this distribution to argue that any axis with an eigenvalue less than that predicted to occur from the appropriate “broken-stick”’ distribution could be ignored as insignificant. Comparison of eigenvalues with those in Frontier’s table of “broken-stick”’ values was undertaken: in all cases discussed in this paper the first four axes proved significant. A second technique, Discriminant Analysis (DSC), was used to test whether pre-defined groups of species visualised on PCA were or were not statistically distinguishable. This multivariate extension of analysis of variance (Marriott 1974) was performed in both a non-stepwise and stepwise manner (all default options, i.e. minimisation of Wilks i, auto F-to-enter and F-to-remove) using SPSS 6.1 (cf. Norusis & SPSS Inc. 1993). 95% confidence limits are used throughout where appropriate. RESULTS The very low number of samples obtained of three taxa (S. anglica, S. devoniensis and S. rupicola) meant that it was not possible to draw firm conclusions relating to any of them. Nevertheless it is worth noting that an initial analysis which included all taxa was useful and that coefficients of variation for characters measured on these species were in the normal range of 10-20%. The four PCA axes which met Frontier’s (1976) criteria had eigenvalues (expressed as percentages of the variance) of 28.4%, 17.9%, 11.7% and 9.3% respectively; there were therefore six biaxial plots of potential significance - which required examination. Examination of these six plots together with initial DSC analysis showed that S. devoniensis was easily and consistently distinguishable from the other species of Sorbus measured on the basis of a combination of: i. character | — its longer petioles (>2 cm + 0-2 cm as opposed to always c. 1-4(-1:9) cm + 0.06); 156 J. PARNELL AND M. NEEDHAM 40 30 20 Number of specimens 10 -3.2 -1.2 0.8 2.8 U4 FiGure 1. Plant scores for Sorbus taxa along Axis 4 (U4) which accounts for 9:4% of the total variance in the initial PCA analysis. Scores for S. devoniensis plants are indicated in black. ii. characters 17 & 18 — the number of teeth on the left and right-hand side of the leaf blade (c. 68 + 12 as opposed to c. 45(-53) + 2); iii. characters 5 & 6 — its more acutely pointed leaves forming an angle of c. 92° + 10° at the apex (as opposed to> 106° + 1-8°); iv. character 8 — the greater distance from the third lateral vein to the base of the leaf (2:2 cm + 0°42 as opposed to 1:7 cm + 0:07); and v. character 7 — the more widely spaced third and fourth lateral veins 1-4 cm + 0-10 as opposed to 0-93 cm + 0-04) for the other species taken together. Fig. 1 shows the distribution of scores along Axis 4 of this initial PCA which, for S. devoniensis is dominated by characters 7, 8 & 15 (the distance between the excursion points of the third and fourth lateral veins at the midrib, the distance from the third lateral vein from the base of the leaf blade to the base of the leaf blade and the number of secondary veins on the leaf blade respectively). Table 2 gives the eigenvector scores for this axis for all characters. Evidently at least these strong scoring characters must be further examined with a more comprehensive data set based on British material as it appears that further biometric work on these taxa will allow sufficiently robust algorithms to be calculated so allowing clear distinctions to be made. Further analysis concentrated on the distinction between S. aria and S. hibernica. As can be seen from Fig. 2 a—~c, PCA offered some support for separation of S. aria from S. hibernica; however it is clear that this support is limited and that considerable overlap of the taxa occurs. In part this is because the plots in Fig. 2 are simple biaxial plots which maximally account for 47% of the variance. A more accurate picture of the separation between these taxa can be obtained by DSC which gave good separation between these two species. Indeed non-stepwise DSC, the more conservative option, gave an overall misclassification rate of only 5-3% and stepwise DSC a highly significant intergroup F-ratio of 26°9 (d.f. 5, 126; p<0-001). Further discussion will be confined to non-stepwise DSC. In general DSC was more successful at correctly classifying S. hibernica (97% success) than S. aria (86% success). The univariate F-ratios for differences between the groups for particular characters highlighted a number of the latter as being of particular differential importance (Table 3). Table 4 lists these characters in decreasing differential order together with their means and 95% confidence limits. As can be seen from these two tables the single most important differential characteristic between MORPHOLOGICAL VARIATION IN IRISH SORBUS L. 157 TABLE 2. EIGENVECTOR SCORES FOR AXIS 4 OF THE INITIAL PCA OF SORBUS TAXA Characters are numbered and abbreviated as in Table | above. Character number and its abbreviation Eigenvector 1. Angbotlf 0-018 2. Angbotrh 0-122 3. Angof3ve — 0-260 4. Angof4ve — 0-262 5. Angtoplf —0:125 6. Angtoprh — 0-084 7. Distbet34 0-472 8. Distbot3v 0-401 9. Distfrm3v —Q-134 10. Distteel —0-148 11. Distteer —0-162 12. Leaflen —0-010 13. Leafwid 0-031 14. Leaflentwp — 0-039 15. Noveins — 0-448 16. Noteeto 0-273 17. Notthlhs 0-201 18. Notthrhs 0-233 19. Petlen — 0-031 20. Petwid — 0-040 S. hibernica and S. aria is the number of secondary veins (character 15); there being an average 22 secondary veins in S. aria and 18 in S. hibernica. As the distance between the third and fourth veins is not significantly different between the two groups (character 7, Table 3) and therefore the vein spacing over the whole leaf is likely to be similar, then it is unsurprising to find that the leaves are significantly smaller (on average over | cm shorter) in S. hibernica than S. aria and that the third vein from the base is closer to the apex in S. hibernica than S. aria. Table 4 also shows that the smaller leaves of S. hibernica are borne on significantly shorter petioles (character 19), have rather more steeply rising veins (characters 3 & 4) but fewer leaf teeth (character 17). So far this analysis has been concerned with differences between individual Sorbus plants and implicitly populations. As S. hibernica is reputedly apomictic it might be expected that the most morphologically similar trees would be in closest physical proximity to one another. Equally if apomixis in S. hibernica is not obligate it might also be expected that a regional analysis would show greater similarity between S. aria and S. hibernica and possibly intermediates where the ranges of the species overlap. The above data were therefore amalgamated on a vice-county basis and vice-county means used to calculate a PCA. Plots of plants against the four significant axes whose eigenvalues exceeded Frontier’s criteria are shown in Fig. 3. As can be seen from Fig. 3 there is almost complete separation between S. aria and S. hibernica in most plots — examination of these makes it clear that this is largely due to the scores for these species on Axis 1. Indeed the separation between S. aria and S. hibernica would be perfect if it were not for a single S. hibernica point. The data from which this point are derived relate to the only collection available of S. hibernica from County Meath. Further examination of this specimen (DBN), which is sterile, shows that there has been doubt expressed about its status, with D. Synnott, the original collector identifying it as S. aria and D. A. Webb as S. hibernica. The specimen is undoubtedly unusual with a mixture of characters of S. hibernica and S. aria — the somewhat upswept leaf-teeth, very dense indumentum, with the leaf veins rising at a steep angle (c. 43°) of S. hibernica, combine with a leaf blade of c. 11 cm, bearing 21 secondary veins, a petiole 1:7 cm long and>51 leaf teeth on the left-hand side of the leaf blade. Quite obviously it would be possible to view this unique specimen as indicating either a hybridisation event linking S. aria and S. hibernica, or as an aberrant member of either species: the available evidence does not easily allow a decision to be made on this question. In the 158 J. PARNELL AND M. NEEDHAM x x 2 x x 5 x aes xX Wax - 3 x RRX x x Xe Xxx Or SE ™ x ox OK By xx * 5 a x x 0 ° ex x ark. ERO x x xX te) ere : % 2 sy i ° maha 5 x Or Qe * 9? x oa xx Xx o Xx o& 1 NS Ns es 1 Bk & x ces Z 2 x 2 x 4 4 ° ° -2.50 -1.25 0.00 1.25 1.5 0.0 15 3.0 A u2 B Us x x 2Tx x ete x . x Xx x x x nee ey Lee 1.25 x xery mK eve) Oo X° oe x X X (0) x xy ta x Bax * x a % KX x Xx kK O& 28a ° 2 % ov ° : : x ~ Xo x° U 0.00 es oe xe x XX x KO x x x x x x 2 xe x x ® U -2 x 1.25 SI ey ‘) 1 co x% xx ., ° -4 2.50 S) x ° ee as em -1.5 0.0 1.5 3.0 -2.50 -1.25 0.00 1.25 Cc U3 D U2 FIGURE 2. PCA plots showing the position of S. aria (o) and S. hibernica ( x ) for various combinations of PCA axes 1-3 (U1-U3). Axis 1 accounts for 29:9%, Axis 2 for 16°6% and Axis 3 for 12°8% of the variance respectively. In Fig. 2 D S. hibernica populations from v.cc. H15, 16 & 17, South-east, West and North-east Galway, are distinguished by (0), all other material is symbolised by (x ). TABLE 3. F-VALUES FOR A ONE-WAY ANOVA FOR PARTICULAR CHARACTERS MEASURED ON SPECIMENS OF S. ARIA AND S. HIBERNICA Character numbers are as in Table | (d.f. = 1, 130). All F-values > 3-84 are significant at p<0-05. F-values significant at p<0-001 are highlighted, by three stars. Character number and its abbreviation WOIDKAHRWNY— . Angbotlf . Angbotrh . Angof3ve Angof4ve Angtoplf . Angtoprh . Distbet34 . Distbot3v . Distfrm3v . Distteel . Distteer . Leaflen . Leafwid . Leaflentwp . Noveins . Noteeto . Notthlhs . Notthrhs . Petlen . Petwid F-value 0-01 2-08 0:36 2°62 0-00 0-00 0-01 7:14 2:45 4-4] 9-28 0-67 TABLE 4. MEAN VALUES FOR CHARACTERS WHOSE UNIVARIATE F-RATIOS INDICATE A SIGNIFICANT DIFFERENCE AT P<0:001 BETWEEN S. ARIA AND S. HIBERNICA TOGETHER WITH MORPHOLOGICAL VARIATION IN IRISH SORBUS L. THEIR 95% CONFIDENCE LIMITS, PRESENTED IN DECREASING ORDER OF F-VALUE Mean + Mean + 95% 95% confidence confidence Character number limits for limits for S. and its abbreviation F-value S. aria hibernica 15. Noveins 58°67 D2 Atta lee 18:-3+0°4 19. Petlen 35:95 1:8+0°1 1-4+0-07 9. Distfrm3v 33:20 8:3 +0°5 6:°7+0°3 6. Angtoprh 29:66 50:0+1°8 56:0+ 1-0 4. Angof4ve 27°61 48:-6+2°5 43-4+0°8 3. Angof3ve 27:03 SUPSar Sill 44:1+ 1-0 12. Leaflen 25:24 9:-9+0°6 8:4+0:3 17. Notthlhs 13-94 50:5 +3°8 43°-5+1°7 3% 3x 1.25 1.25 x %&, en x ™m X 5% x 0.00 ols ix 0.00 x a U xX x x ° ajed-25 ° ° TAOS oy Ole 2 ° SUAS 2% 2x + Ft —<$—$}+—_—__}_—_ -1.25 1.25 44) 0 1 A U2 B U3 3% 1 x 1.25 2.50 7 a x x 6% 0.00 a ab: 1.25 x x ex = % OWide U 0.00 XR KS W425 fe] fo) x x x x 4 ° 4X 0 Xo ° -1.25 ° oe 2% 3% nt Se a Pe Gene Et -2 -1 0 1 -1 0 | U4 U3 C D FiGur_ 3. PCA plots of Axes 1-4 (UI—U4) for an analysis of individuals of Sorbus (Sorbus aria (0) and Sorbus hibernica (x ). Axis 1 accounts for 32°6%, Axis 2 for 17:2%, Axis 3 for 15-1% and Axis 4 for 10-1% of the total variance. Material from v.c. H2 North Kerry, from v.c. H22 Meath and from v.c. H26 East Mayo are indicated by the numbers 1-3 respectively placed to the immediate left of the appropriate symbol. 160 J. PARNELL AND M. NEEDHAM TABLE 5. MEAN VALUES FOR CHARACTERS WHOSE UNIVARIATE F-VALUES INDICATE SIGNIFICANT DIFFERENCES AT P <0:001 BETWEEN VICE-COUNTY MEAN VALUES FOR S. ARIA AND S. HIBERNICA TOGETHER WITH THEIR 95% CONFIDENCE LIMITS, PRESENTED IN DECREASING ORDER OF F-VALUE Mean + Mean + 95% 95% confidence confidence Character number limits for limits for S. and its abbreviation F-value S. aria hibernica 15. Noveins eS Messe le 18:3+0°6 17. Notthlhs 34-9] Dore ae 1/93) ASD ataaltea 19. Petlen 32:01 1:9+0:2 1:4+0:1 9. Distfrm3v 28:02 oe Slar OPS) 6:6+0°3 12. Leaflen 23:06 10:0 + 0:6 8:3+0°3 13. Leafwid 17:97 6:5+0-4 3*65n022 18. Notthrhs 9-64 53°64: 7-0 45:-4+2°4 circumstances it may be most useful to accept that this specimen is not determinable at present and that therefore, from the perspective of describing the core characteristics of the species, it is best to lay it to one side. In fact by trial and error it was found that removal of this single point had very little effect on the scatter diagrams produced by PCA and those produced where the specimen had been removed are therefore not reproduced here. DSC analysis of the data, after removal of the aberrant Meath specimen, produced a similar result to that seen before, though naturally less importance should be attached to the values obtained through this analysis as they are based on vice-county means and exclude the, perhaps critical, Meath specimen (Table 5). Though Table 5 indicates that this DSC gave results broadly similar to the previous analysis shown in Table 4, it is of interest that the new analysis indicates that leaf width is also a taxonomically useful feature enabling distinction to be made between S. aria and S. hibernica and that the angles that the secondary veins make with the midrib are less diagnostically important. Further examination of Figs 2 D & 3 showed no evidence whatsoever for a closer morphometric linkage in the variation pattern of S. hibernica within a vice-county or for adjacent vice-counties than between geographically remote vice-counties (e.g. note Kerry, Mayo and Meath (nos 1-3 in Fig. 2) are obviously well separated from the rest of the S. hibernica records from their province and the wide spread of points in Fig. 2 D for material from Galway). CONCLUSIONS The above data analyses clearly show that it is possible to use a range of characters to distinguish S. aria from S. hibernica and that S. devoniensis is relatively easily distinguished from other Sorbus species in Ireland. The analyses have indicated a number of extra differential morphological characters, which are particularly useful for distinguishing sterile material. In particular it is clear that the number of teeth on the leaf, the angle of the secondary veins with the midrib and the length of the petiole are useful differential characteristics enabling S. aria to be distinguished from S. hibernica. It is clear that there are still difficulties associated with differentiation of material but these new characters allow most specimens to be determined without much error or difficulty. The difficulty experienced in relation to assignment of specimens to either S. aria or S. hibernica is of relevance to the question of the level, if any, of outcrossing in the polyploid microspecies discussed by Proctor, Proctor & Groenhof (1989). It is clear that there are examples listed by Richards (1975) of apparent hybridisation between S. aria and various polyploid microspecies and it may well be that the difficulties faced in this work have arisen in part due to a rare hybridisation event(s). One of the surprises of this work was the similarity in relative variability of the two species, or occasionally the greater degree of relative variability in S. hibernica as compared to S. aria. For MORPHOLOGICAL VARIATION IN IRISH SORBUS L. 161 example the percentage coefficient of variation is 12% and 13% respectively for the number of veins in S. aria and S. hibernica and the corresponding figures are 18% and 24% respectively for petiole length. This seems to indicate that the assumption in the literature that S. aria is a relatively variable species is false, at least in Ireland. Experimental investigation by us of the breeding system of S. hibernica has so far proved inconclusive, but if outbreeding does occasionally occur in S. hibernica it may go some way towards explaining the relatively high degree of variability in that species. Webb & Scannell (1983) speculate on the origin of some of the S. aria material in East Connaught (East of Galway), which is the main centre of distribution of this species in Ireland. They draw attention to the fact that some of this material may be derived from introductions or plantings. However, after considerable discussion they accept that the species is native. Scannell & Synnott (1987) draw attention to the fact that material in v.c. H21 (Co. Dublin) is probably introduced. However of the four localities cited (Doogue ef al. (unpublished)), two are in hedgerows, are probably bird-sown and are not clearly non-native. Indeed none of the S. aria specimens in our survey appear to have been clearly planted or derived from planted material; however, the possibility remains that much Irish material has been derived from a relatively few introductions, which in turn could explain the relatively low variability of this species in Ireland. Evidently further work is needed on S. hibernica to confirm its apomictic nature and also to look more closely at its relationship to other Irish and, eventually, British material. DNA sequencing, which we intend to commence soon, is likely to be able to resolve these difficulties but such work, interesting though it may well be, is not going to alter the difficulties experienced by field-workers: therefore more biometric work on this complex is required. REFERENCES CLAPHAM, A. R., TUTIN, T. G. & Moore, D. M. (1987). Flora of the British Isles. Cambridge University Press, Cambridge. DooGuE, D., NASH, D., PARNELL, J., REYNOLDS, S. & WySE JACKSON, P. (1998). Flora of County Dublin. Dublin Naturalists Field Club. (Manuscript in proof). FRONTIER, S. (1976). Etude de la décroissance des valeurs propres dans une analyse en composantes principales: comparison avec le modeéle du baton brisé. Journal of experimental marine biology and ecology 25: 67-75. JACKSON, D. A. (1993). Stopping rules in principal components analysis: a comparison of heuristical and statistical approaches. Ecology 74: 2204-2214. LEGENDRE, L. & LEGENDRE, P. (1983). Numerical ecology. Elsevier, Amsterdam. Marriott, F. H.C. (1974). The interpretation of multiple observations. Academic Press, London. Norusis, M. J. & SPSS INc. (1993). SPSS for Windows. Professional Statistics. Release 6.0. SPSS, Illinois. Proctor, M.C. F., Proctor, M. E. & GROENHOF, A. C. (1989). Evidence from peroxidase polymorphism on the taxonomy and reproduction of some Sorbus populations in south-west England. New phytologist 112: 569- SWS: RICHARDS, A. J. (1975). Sorbus L., in STACE, C. A. ed. Hybridisation and the Flora of the British Isles, pp. 233— 238. Academic Press, London. SCANNELL, M. J. P. & SyNNotTrT, D. (1987). Census catalogue of the flora of Ireland. Stationery Office, Dublin. STACE, C. A. (1997). New Flora of the British Isles, 2nd ed. Cambridge University Press, Cambridge. Wess, D. A., PARNELL, J. & DooGueE, D. (1996). An Irish Flora, 7th ed. Dundalgan Press, Dundalk. Wess, D. A. & SCANNELL, M. J. P. (1983). Flora of Connemara and the Burren. Cambridge University Press, Cambridge. (Accepted January 1998) kektany idk rinse wT Mesa ee id ne Nis “idly weno 3 ra ti ponnats wh “ye waite: & Rte P is ¥ ‘| anne 10 penne ree ait rats nv abe ; ary TeORITEE’ @ Henney ie otiaew at oIsede” ate , vowort L basubertht Vidar: any et (atl idud 8 a ut fairy 2, i v oi I ve sts heaters qidading, ¢ we eS aK voragt york at TG, vit (Gai ye i ane sient t ysirnue qe ai ehoriseqe, ny, 2G OHIO | ti) . ty. j . ay my ey) Ure } A Geiit Woe eh atherte? ailiehianc kala TAS at leita seg ; 5 Sy ita SH nisin» bias f fen it i shit We at age © wel : i Thi hac | + 4 Lhe i \ Fi Youlut oele Ite alas ee nit poate Oh wy sium mt 2. ne! pokeon a es Winey .brn dell vaio OF quer ‘ . . n y tw & LED bs 13 4% AG, VES" oes 2 a t > . f Pe seh ye ee xt i 6 ig P ren . 02 3 ii ; ay e9- vert Mremaat + cmrtts ee hd hs Vt ty ng oF i 0 PATS AW TPG Peat, TON FE) QO Sr te eA th rj ys ay i2 mi ‘ ‘4 ; omy ‘ ditt ae ‘ . £ -~ ridin bak pater, Taha es cane ne ah aie = rE w. ? pune oy ier a beige ithe ph ech = eee, tae ig ou io PAR & Re: ease ae 2. aS f : af ) 44 & - 2 haake t os eal CAa 3% ~ * ; ; ' “ j , 1 A ™ ‘ * y | > nie ete % y ag ec Ce ah uke hs | *, “ye ‘ "th ‘ b A ays E $ au if Ww , ai h x ’ 7 hy i if ’ Re i \ oes Py wel & ‘ yh) } ; : . te et . »' * rh i TH aad ‘ +k fi ee ’ bY 4 b-y* | i hy ¥ ., ‘ > a t 2} ‘ ~ q i‘? 4 4 ~ ote i vw § ¥ Ay e ‘ + - > 4 4 # . n , ; i tet att F 5 5 4 1 “a « ‘ ty a ; mi . h - ” . Hi Se es > bie — Watsonia 22: 163-172 (1998) 163 J. E. Dandy & G. Taylor’s unpublished study of Potamogeton x sudermanicus Hagstr. in Britain, with an account of the current distribution of the hybrid C. D. PRESTON I.T.E., Monks Wood, Abbots Ripton, Huntingdon, Cambs., PEI7 2LS and D. A. PEARMAN The Old Rectory, Frome St Quintin, Dorchester, DT2 OHF ABSTRACT In 1942 J. E. Dandy & G. Taylor prepared an account of Potamogeton x sudermanicus (P. acutifolius Link x P. berchtoldii Fieber), a hybrid they reported as new to the British Isles on the basis of specimens collected near Wareham in Dorset (v.c. 9). The paper was intended for the Journal of botany but the journal ceased publication before it could appear. The text of the paper is published here for the first time. The hybrid was first collected in the Wareham area by J. H. Salter in 1920 or 1921. Recent surveys show that the hybrid is more frequent than P. acutifolius and grows in the absence of P. berchtoldit. KEYWORDS: Potamogeton acutifolius, P. berchtoldii, Potamogetonaceae, vegetative reproduction. INTRODUCTION Our current understanding of the genus Potamogeton in the British Isles rests on foundations laid by J. E. Dandy and G. Taylor, colleagues at the Natural History Museum, London, who collaborated closely on taxonomic studies of the genus from the mid 1930s until Taylor left the Museum for Kew in 1956. One result of their work was a series of papers in the Journal of botany entitled ‘Studies of British Potamogetons”’’, 18 of which were published between 1938 and 1942. In our view, these papers represent “‘one of the most critical and scholarly contributions made this century to the taxonomic study of the British flora” (Preston 1995a). This series of papers stopped abruptly when the Journal of botany ceased publication in 1942. In 1985 C.D.P. visited Sir George Taylor, who generously lent him a copy of the manuscript monograph The British species of Potamogeton L., which Dandy & Taylor had prepared but never published, together with other associated papers. Included amongst the latter was the completed typescript of a 19th paper in the “Studies of British Potamogetons”’ series. This paper covers Potamogeton x sudermanicus Hagstr., the hybrid between P. acutifolius Link and P. berchtoldii Fieber, and was virtually ready for publication when the Journal of botany folded. The hybrid had been detected by Dandy & Taylor from herbarium specimens collected near Wareham in Dorset (v.c. 9). It still survives in this locality, but has not been found elsewhere in Britain. The record of P. x sudermanicus was published by Good (1948) and included in Dandy’s (1975) treatment of the British hybrids. In this paper we publish the text of Dandy & Taylor’s paper, which still merits publication for its treatment of the history and taxonomy of P. x sudermanicus in Britain (a striking illustration of the long shelf life of taxonomic publications). It is also of historical interest as a fine example of the work of Dandy & Taylor. Finally, the belated publication of this paper is an appropriate way to mark the 60th anniversary of the start of the “Studies of British Potamogetons”’ series in 1938. We have ourselves been interested in the hybrid P. x sudermanicus for some years, making repeated 164 C. D. PRESTON AND D. A. PEARMAN visits to the Wareham area to study its distribution in relation to that of its parents. We therefore follow Dandy & Taylor’s paper with an updated account of the distribution of the hybrid in Dorset. DANDY & TAYLOR'S UNPUBLISHED STUDY THE TYPESCRIPT The paper on P. x sudermanicus was initially numbered “Studies of British Potamogetons. — XVII’, showing that it was written before the publication of Studies XVII and XVIII. In the event Studies XVII and XVIII were taken up with papers furthering the controversy between Dandy & Taylor and Professor J. W. Heslop Harrison, which appeared in the issue of the Journal of botany intended for publication in July 1942 (although not published until April 1944). Although the typescript is not explicitly dated, there is strong internal evidence that the copy in Taylor’s possession was typed between June and October 1942. It discusses the collections of P. x sudermanicus made by F. C. Steward in June 1942, but details of his collections made in October 1942 are added in handwritten annotations. Further evidence comes from the fact that the typed phrase “This year...” has been altered twice by hand, initially to “Last summer...” and then to “In 1942...”. There is little doubt that the collapse of the Journal of botany was the main reason why Dandy & Taylor’s 19th Study was never published. The Journal of botany had become closely associated with the Botany Department of the British Museum (Natural History) during the long period (1880-1924) when it was edited by James Britten, and it came to serve in place of an official periodical (Stearn 1981). John Ramsbottom, the Museum’s Keeper of Botany, took over the editorship in 1938. The burden of running the Department of Botany in wartime must have been considerable, but in 1985 Taylor still blamed Ramsbottom’s laziness for the collapse of the Journal. Ramsbottom had many gifts but his fellow mycologist G. C. Ainsworth (1986) noted that “to exasperate people by procrastination was an integral part of his character’. Letters from Taylor to Dandy now held in the manuscript collection of the Natural History Museum (DF440/63) show that Taylor pressed Dandy to publish the note on P. x sudermanicus after 1942, initially with papers on P. x suecicus published in the Transactions of the Botanical Society of Edinburgh (Bance 1946; Dandy & Taylor 1946). On 20 May 1945 Taylor wrote to Dandy and after discussing the P. x suecicus papers added “We might also put in the paper on sudermanicus if you can lay hands on it”. On 3 July 1945 he reverted to this suggestion: ““There should be no difficulty in getting the swecicus papers to [H.R.] Fletcher by the end of September. Do you think that we could plug in another paper of Pot records at the same time? He seems quite keen to have them and here is a grand opportunity for a grand slam. What about the sudermanicus effort as well? I do realize how much work will be required to prepare all these for publication but if you can let me know how I can help I shall gladly do so.” Both P. x suecicus papers were read by title at a meeting of the Botanical Society of Edinburgh on 13 June 1946 and subsequently published. Another paper by Dandy & Taylor, “New and interesting British records of Potamogeton”’, was read by title on 15 May 1947, usually a formal prelude to publication, but it never appeared in print. On 13 June 1947 Taylor forwarded to Dandy some photographs of the leaves of P. acutifolius and P. x sudermanicus taken by Miss H. M. Bance, who had undertaken anatomical studies of P. x suecicus at Taylor’s behest, but he clearly failed to prompt Dandy into the work needed to complete the paper on this hybrid. The typescript of the unpublished study is marked up with corrections in Dandy’s hand and some more tentative comments by Taylor. Our aim is to print it as Dandy & Taylor intended to publish it, incorporating the corrections made by Dandy. We have incorporated minor rewording without comment, but have drawn attention to some more significant changes or annotations as numbered notes. These changes show, for example, how the austere and authoritative style of the Studies was achieved by the ruthless excision of material of a more speculative nature. We have also added some other explanatory notes. We have retained the typographical conventions of the 1940s, e.g. the placement of the hybrid sign x before the generic name, to maintain conformity with the other papers in the series. Sir George Taylor died in 1993 and his books and papers were bequeathed to the National Library of Scotland (accession no. 9533). The manuscript published here was presumably amongst them, but Taylor’s manuscripts have not yet been catalogued and in any event all manuscripts held by the National Library are currently unavailable because of building work. POTAMOGETON x SUDERMANICUS IN BRITAIN 165 TEXT OF THE STUDY The text of the study is as follows: STUDIES OF BRITISH POTAMOGETONS. — XIX. BY J. E. DANby, M.A., AND G. TAYLor, D.Sc. XIX. x POTAMOGETON SUDERMANICUS IN BRITAIN. Hybrids between the “‘pusilloid” (linear-leaved) species of Potamogeton are remarkably rare when we consider the comparative frequency with which some of the broad-leaved species interbreed’. It is true that Hagstrom in his ‘Critical Researches’ recognized eleven hybrid combination of “pusilloid” species in Europe, but some of the plants which he treated as hybrids are without doubt only states of species; and of the British plants identified in print as “‘pusilloid” hybrids by Hagstrém, A. Bennett, and others not one is of hybrid origin. Nevertheless genuine hybrids between “‘pusilloid”’ species do occasionally occur in Britain, as elsewhere, and one of them forms the subject of the present note”. In working through the pondweeds of the Druce Herbarium, Oxford, we came across an unidentified specimen from the Wareham district of Dorset which might at first sight have passed for a broad-leaved state of P. Berchtoldii but for the obvious strong compression of the stem and the presence of a fruiting- carpel of too large size. On examining the specimen more closely we found that the leaves have, besides the usual three vascular nerves, an irregular number of fine (often broken) sclerenchymatous nerves of the type which characterizes the leaves of P. acutifolius and P. compressus. The presence of these nerves, in association with a strongly compressed stem, at once suggested a hybrid of P. Berchtoldii with P. acutifolius, which occurs in the Wareham district whereas P. compressus does not. Further investigation of other characters showed the plant to be in all respects intermediate between P. acutifolius and P. Berchtoldii. For example, the fruiting-carpel (the only one developed, and possibly not fertile, is midway in size and form between the two species. The stipular sheaths are open and convolute as in both P. acutifolius and P. Berchtoldii. Thus the morphological evidence convinces us that the plant is a hybrid between these species, a conclusion which is supported by the fact that both have been collected in the neighbourhood of Wareham. The plant under discussion was collected by Mr. A. W. Graveson about 1927 in the ditches of the water-meadows near Redcliff Farm*, south-east of Wareham. In 1942, hearing that Dr. F. C. Steward intended to visit Dorset in June, we requested him to search for the plant: on being given details of the locality he succeeded in collecting a good series of specimens, though unfortunately these were not in flower. In October he paid another visit to the place and obtained further specimens, again sterile but this time bearing winter-buds intermediate between those of the parent species’. Highly interesting as it is, this hybrid between P. acutifolius and P. Berchtoldii is not new to science. It was described from Sweden by Hagstrém in his ‘Critical Researches’, p. 73, fig. 28 A—E, under the name x P. sudermanicus and with the formula P. acutifolius x pusillus (his ““P. pusillus” being P. Berchtoldii). Hagstrom’s description and figures agree well with the Dorset plant except that the peduncles of the Swedish plant are given as only 6-8 mm. long as against 24 mm. in Mr. Graveson’s specimen, while the spikes are described as entirely barren whereas, as we have already mentioned, a single fruiting-carpel is present in the material from Dorset. This solitary fruiting-carpel may, however, be a chance development such as we have observed occasionally in such sterile hybrids as x P. fluitans, x P. nitens, and x P. sparganifolius’. The longer peduncles of the Dorset plant is also without significance as its length comes well within the limits of variation to be expected from the character of the parent species. x P. sudermanicus was named after Sudermania, in Sweden, where the type was collected in 1831. The type-locality, as Hagstrom remarked in describing the hybrid, is at the most northern border of the distribution area of P. acutifolius. He added that the plant probably belongs to the greatest rarities of the © vegetable kingdom, and this may well be true despite the discovery of the Dorset station’. Mr. Graveson’s specimen apparently ranks as the first authentic record of x P. sudermanicus from the British Isles, an earlier record for East Sussex being an error. The Sussex plant, which was collected near Camber Castle, Icklesham, by C. E. Salmon in 1900, is quite normal P. pusillus, and its 166 C. D. PRESTON AND D. A. PEARMAN treatment as x P. sudermanicus by Bennett in Journ. Bot. Ix. 55 (1922) is inexplicable as its stipular sheaths are of course tubular, whereas those of x P. sudermanicus and both its parent species are open. This erroneous record of x P. sudermanicus has been repeated in other works including the ‘London Catalogue’, Ed. 11 (1925), p. 46; Druce’s ‘British Plant List’, Ed. 2 (1928), p. 117; and Wolley-Dod’s ‘Flora of Sussex’ (1937), p. 465. It was referred by us to P. pusillus in our note on that species (Journ. Bot. Ixxviii. 5) in 1940. Following is the brief synonymy and distribution of x P. sudermanicus as a British plant. P. ACUTIFOLIUS xX BERCHTOLDII = x P. SUDERMANICUS Hagstr. Crit. Res. 73, fig. 28 A-E (1916)'°. P. acutifolius x pusillus Hagstr. op. cit. 73 (1916). We have seen specimens from only one vice-county:— (9) DORSET. Ditches in water-meadows near Redcliff Farm, Arne, c.1927, A. W. Graveson, Ref. 5 (Herb. Druce); June and Oct. 1942, F. C. Steward (Herb. Brit. Mus.). NOTES 1. The sentence “The reason may well be that the “pusilloid” species depend more on vegetative winter-buds than on seeds for their reproduction.” and Taylor’s addition “and many are shy flowerers”’ follow this sentence but have been enclosed in square brackets, apparently to denote that they should be deleted. The following paragraph, in Taylor’s handwriting, is attached to the copy of the P. x sudermanicus typescript and also deals with this issue; it is not clear where, if anywhere, Taylor intended it to be inserted. {It is not] easy to understand why there should be apparent antipathy between closely allied species which grow in close association and in such circumstances frequently produce an abundance of fruit. Whether these fruits are viable or not is a matter for experiment or observation in the field but, in addition to possible increase by germination, the pusilloid species always provide for perpetuation by vegetative means: Fernald, indeed, has suggested that winter buds “are the usual, if not the only, means of reproduction’. As the plants mature, whether they have fruited or not, they invariably produce winter-buds. In many situations pusilloid species are shy in[?] flowering and depend on vegetative propagules for their survival.” The first words in the paragraph are illegible on my photocopy and the material in square brackets is my interpolation. Taylor’s quotation is taken from Fernald (1932, p. 21). 2. The other British pusilloid hybrids are P. acutifolius x friesii, described as P. x pseudofriesii by Dandy & Taylor (1957), and P. pusillus x trichoides, described as P. x grovesii by Dandy & Taylor in Sell (1967). 3. Dandy had marked this sentence by a line in the margin and added a question mark. 4. The spelling Redcliff was the norm when Dandy & Taylor wrote, although Redcliffe appears on modern maps. 5. The phrase “In 1942” originally appeared as “This year’ and then as “‘Last summer’’. The phrase “On being given details of the locality...” replaces “He responded with enthusiasm, and having obtained details of the locality from Mr. Graveson...’’. 6. This sentence is a manuscript addition in Dandy’s hand. 7. The first half of this and the second half of the preceding sentence have been marked with a double line in the margin and Taylor has commented “*? Simpson’s specimen’’. This presumably refers to a specimen of P. x sudermanicus collected by N. D. Simpson on 27 June 1945 and now in BM which also bears an enlarged carpel. The occasional development of swollen carpels resembling immature fruits is a feature of the Wareham population of P. x sudermanicus (Preston 1995b). 8. For a summary of subsequent records of P. x sudermanicus in Europe, see Ploeg (1987). The hybrid has been recorded from England, the Netherlands and Sweden; a more tentative report from Germany requires confirmation. 9. In 1973 Dandy determined as P. x sudermanicus a specimen collected as P. acutifolius from Stoborough Meads by J. H. Salter in 1920 or 1921 (NMW), which predates Graveson’s specimen. 10. The reference is to Hagstrém (1916). POTAMOGETON x SUDERMANICUS IN BRITAIN 167 MORPHOLOGY OF P. X SUDERMANICUS A description and illustration of P. x sudermanicus, based on studies of living plants and herbarium material from the Wareham population, has already been published (Preston 1995b) and need not be repeated here. Two aspects can, however, be elaborated. In studies of the living plant it became clear that the compression of the stems of the hybrid forms a useful quantitative character to distinguish it from both P. acutifolius (which has strongly compressed to flattened stems) and P. berchtoldii (which has terete to slightly compressed stems). This conclusion was based on measurements of the longest and shortest axes of the stems of all three species in cross section. These data are summarised in Table 1 and illustrated in Fig. 1. The hybrid is also intermediate in floral characters between P. acutifolius (which usually has one carpel per flower) and P. berchtoldii (which usually has 4-5), as demonstrated by the data in Table 2. P. compressus, which is closely related to P. acutifolius, usually has two carpels TABLE 1. COMPRESSION OF STEMS OF POTAMOGETON ACUTIFOLIUS, P. BERCHTOLDIT AND THEIR HYBRID P. x SUDERMANICUS Stem compression Taxon Number of stems examined Mean Range P. acutifolius 25 3°38 2:°8-4:7 P. x sudermanicus 43 2:0 1:6—2°4 P. berchtoldii 32 1-4 1-0-1-6 Stem compression is the ratio of the longest to the shortest axis of the stem in cross-section, measured on fresh plants. Based on material of P. acutifolius from Dorset, Norfolk and Sussex, P. berchtoldii from a range of sites in the British Isles and P. x sudermanicus from Dorset. 25 = Potamogeton acutifolius O P. X sudermanicus 4 Potamogeton berchtoldii a x< © | Q 15 @ |: 0) | = & 0.5 Ors 4084 a0. Sch 2106.0 OF eeay O88, >) 10:95 a0 0 shortest axis FiGuRE. |. Stem compression of P. acutifolius, P. berchtoldii and their hybrid P. x sudermanicus. The dimensions — of the longest and shortest axes of the fresh stems (mm) were measured in cross sections of fresh material of P. acutifolius from Dorset, Norfolk and Sussex, P. berchtoldii from a range of sites in the British Isles and P. x sudermanicus from Dorset. Regression lines are shown for P. acutifolius (r° = 0-26), P. x sudermanicus (1° = 0:78) and P. berchtoldii (t7 = 0-77). 168 C.D. PRESTON AND D. A. PEARMAN TABLE 2. NUMBER OF CARPELS IN THE FLOWERS OF POTAMOGETON ACUTIFOLIUS, P. BERCHTOLDIT AND THEIR HYBRID P. x SUDERMANICUS Number of carpels per flower Number of flowers Taxon examined | 22 3 4 5) 6 i P. acutifolius 19 19 0) 0 0 0 0 0 P. x sudermanicus 59 2 36 10 | 0 0 0 P. berchtoldii 89 0 0 6 64 16 2 1 Based on material of P. acutifolius from Dorset and Norfolk, P. berchtoldii from a range of sites in the British Isles and P. x sudermanicus from Dorset. The data for P. x sudermanicus exclude one malformed flower with four carpels and eight stamens. per flower, though a minority of flowers have one (Charlton & Posluszny 1991). The presence in P. x sudermanicus of a substantial minority of flowers with a single carpel perhaps provides morphological evidence for the fact that P. acutifolius rather than P. compressus is one parent, and thus supports a conclusion reached by Dandy & Taylor on geographical grounds alone. DISTRIBUTION OF P. X SUDERMANICUS AND ITS PARENTS IN DORSET All records of P. x sudermanicus in Britain are from ditches in grazing marshes by the River Frome south of Wareham. The earliest known specimen was collected from “Stoborough Meads, Wareham” by J. H. Salter (NMW). Salter’s natural history diaries, now in the National Library of Wales at Aberystwyth (NLW MS 14444B), throw some light on the discovery of the hybrid. Salter visited Wareham on 6 September 1920, “with a special view to the old marsh ditches’. On leaving Wareham, he “took the path alongside the river for Redcliff. Close to Redcliff Farm I turned into the Stoborough Meads and at once found the long-looked-for Potamogeton acutifolius in great plenty”. On 9 August 1921 he returned “by first train to Wareham to have a good hunt for water plants in the old ditches’’. After visiting ditches north of the town he made for Ridge. “As I followed the path alongside the riverside to Redcliff I saw plenty more Juncus obtusifolius in the overgrown ditch on the right. I diverged into the meadows, but this time saw only a little Potamogeton acutifolius and that not in flower or fruit.”’ Salter’s specimen, labelled “P. acutifolius” but actually P. x sudermanicus, is dated 1921. It is, however, a flowering specimen which suggests that he may have collected it in 1920, the year he first saw the plant. A. W. Graveson collected specimens of P. x sudermanicus from “ditches of Stoborough water meadows near Redcliff’ on 14 September 1928 (OXF, fide Dandy’s card index at BM). The fact that both Salter and Graveson unwittingly collected P. x sudermanicus near Redcliffe Farm in the 1920s suggests that it was already well established by then. F. C. Steward’s specimens collected in June and October 1942 are labelled “‘drainage-ditch in water-meadows near Redcliff Farm, Ame”. N. D. Simpson collected the hybrid from shallow ditches at ““Stoborough Meadows” on 27 June and 31 October 1945, A. H. G. Alston gathered it “near Redcliff Farm” on 26 May 1946 and Taylor made a copious collection “near Redcliff Farm” on 17 August 1946; all these collections are in BM but duplicates of Alston’s and Taylor’s were distributed to numerous other herbaria. All these specimens may have been collected from the same area. A map on the reverse side of N. D. Simpson’s collecting stub for collection 45050 (BM) shows that he gathered P. x sudermanicus on 27 June 1945 from the straight ditch running west from Redcliffe Farm; he had collected P. acutifolius on 5 June 1945 from another ditch in this area west of the Farm. Writing to J. E. Dandy on 2 August 1945, Mrs W. B. Watt also described a site for P. x sudermanicus which must lie west of the Farm and south of the River Frome. This area is also shown on a map tipped into the P. x sudermanicus pages of G. Taylor’s typescript of the monograph of Potamogeton he drafted with Dandy. It seems likely that all the early collections of the hybrid were made in the area west of Redcliffe Farm, where it still grows with P. acutifolius (Fig. 2). POTAMOGETON x SUDERMANICUS IN BRITAIN 169 oN toborough aN \ NY: / LONG 24 m- \ How * apes | INN _”~ Crown Copyright 1997 92 93 94 FiGuRE. 2. The distribution of P. acutifolius, P. berchtoldii and their hybrid P. x sudermanicus in the area south- east of Wareham, as recorded between 1987 and 1997. Ditches in which P. acutifolius has been recorded in the absence of P. x sudermanicus are shown by short dashes (- - - - - - - ), those in which P. acutifolius and P. x sudermanicus have been recorded are shown by a hatched line (+~++++++-) and those in which P. x sudermanicus has been recorded in the absence of P. acutifolius are shown by long dashes (— — —). The only locality for P. berchtoldii is denoted by a star (4%). The grid lines are | km apart and the relevant eastings and northings are shown; the entire area lies within 10-km square SY/9.8. Fieldwork between 1987 and 1997 has established that P. x sudermanicus can usually be found both west and east of Redcliffe Farm south of the River Frome (Fig. 2). It grows in water 25-75 cm deep and of pH 7-8 in grazing marsh ditches. Like many aquatics, it may vary in quantity from year-to-year and it is sometimes abundant, especially in the meadows east of Redcliffe Farm. It may be the most numerous macrophyte in some lengths of ditch, either occurring as dense masses in well-vegetated — areas or as scattered plants on the rather open substrate of ditches cleared out the previous year. Interestingly, specimens collected by N. D. Simpson on 31 October 1945 were “growing up out of the mud after partial clearing of the ditches”. P. x sudermanicus may grow in smaller quantity, sometimes as scattered plants amongst dense populations of Elodea canadensis, and it persists in shaded water 170 C. D. PRESTON AND D. A. PEARMAN amongst dense stands of Phragmites australis. It often grows with Callitriche sp., Eleogiton fluitans, Elodea canadensis, E. nuttallii, Lemna minor, L. trisulca, Potamogeton natans and Spirodela polyrhiza; less frequent associates include Alisma plantago-aquatica, Glyceria fluitans, Hydrocotyle vulgaris, the aquatic variant of Juncus bulbosus, Mentha aquatica, Persicaria amphibia, Potamogeton acutifolius, P. pectinatus, Ranunculus flammula, Sparganium emersum and the moss Fontinalis antipyretica. In 1996 Bryan Pickess reported a narrow-leaved pondweed from ditches north of the River Frome. - Exploration of the site, an R.S.P.B. reserve, revealed that the plant was P. x sudermanicus and that it was widespread but not abundant in this area. This represents a new site for the hybrid, adjacent to the known sites but separated from them by the river. The ditches north of the Frome have a limited aquatic flora, and in many places they are almost choked by Phragmites australis. In 1997 P. x sudermanicus grew in the more open areas of the Phragmites-lined ditches, with only a few associates including Elodea nuttallii, Lemna minor, L. minuta, L. trisulca and Persicaria amphibia. The putative parent P. acutifolius has a much more restricted distribution than P. x sudermanicus in the area (Fig. 2). We have consistently found it only in ditches west of Redcliffe Farm; repeated surveys of the ditches east of the Farm have failed to reveal it there. P. acutifolius formerly had a more widespread distribution in the Wareham area. Nineteenth century records indicate that it grew north of Wareham, between the railway station and the town, as well as in the area where it still occurs near Redcliffe Farm, south of the town. The meadows by the River Piddle north of Wareham still have a rich aquatic flora (despite the fact that they have been bisected by a major road) but P. acutifolius has not been seen here since 1921. Unfortunately, the historical records from the area near Redcliffe Farm are too imprecise to indicate the particular meadows in which the species grew, and we do not know whether it was formerly found east as well as west of Redcliffe Farm. The Wareham localities are the only ones recorded in Dorset for Potamogeton acutifolius, which is a rare species in Britain (Preston & Croft 1997). P. berchtoldii is more common in Dorset and nationally than P. acutifolius. It has been recorded from the vicinity of the P. x sudermanicus sites in the past. It was collected from a “stream on the way to Ridge Farm, Wareham”’ in 1917 (BM) and from a “‘pool near Ridge” in 1934 (BM). D. A. Cadbury gathered specimens of P. acutifolius and P. berchtoldii in 1959 which are both labelled ‘‘ditch by R. Frome above Stoborough” but it is not clear whether they were growing in the same ditch. P. berchtoldii has not been collected subsequently in the immediate vicinity of P. acutifolius or P. x sudermanicus, but it still persists in several places north of Wareham. DISCUSSION The presence of Potamogeton x sudermanicus in the Wareham area since 1920 or 1921, when it was first collected by J. H. Salter, is a striking example of the local persistence of Potamogeton hybrids. It is particularly noteworthy that it has persisted in grazing marsh ditches. In this respect its history provides a contrast with P. x lanceolatus Sm., the hybrid between P. coloratus Hornem. and P. berchtoldii. This hybrid failed to persist in a ditch in Cambridgeshire where it was recorded once, in 1880, but has survived since the 19th century in three streams in western Ireland. Preston (1993) argued that as a sterile hybrid it might be more likely to persist in the open habitats provided along the beds of a stream or small river than in a ditch, where it could be eliminated by competition from emergent species during phases when the ditch becomes overgrown. Postcards of Wareham taken from Redcliffe which were published by the Dorset firm of Delpool in 1984 and 1987 (numbers WM-3 and WM-3(R)) show that the ditches in which P. x sudermanicus grows north of the river Frome and which are now narrow and colonised by abundant Phragmites australis have been broad and open in the past. The Dorset population of P. x sudermanicus is presumably dependent for its survival on the occasional clearing of the ditches in which it grows. Another noteworthy feature of the distribution of the hybrid is that for much of its Dorset range it is present in the absence of both parents. It seems likely that the hybrid arose in a ditch where both parents grew at Wareham, rather than arriving by long-distance dispersal of pollen or seed. P. berchtoldii formerly grew in the Redcliffe Farm area although it has not been found there in recent years. It is possible that P. x sudermanicus first arose in the area where it still grows with P. acutifolius west of POTAMOGETON x SUDERMANICUS IN BRITAIN 171 Redcliffe Farm and attained its current distribution by spreading to the ditches east of the Farm and north of the River Frome. Another possibility is that P. acutifolius formerly grew in the ditches now occupied by the hybrid alone, but has contracted in distribution. In this case P. x sudermanicus might have arisen anywhere in this area, or indeed in more than one locality. Evidence from molecular studies, such as that obtained for the other rare Dorset Potamogeton hybrid, P. x schreberi, by Hollingsworth et al. (1995), might establish the relationship of P. x sudermanicus to the P. acutifolius plants currently present at Wareham and determine whether the hybrid is represented by one or more clones. Although it is clear that P. x sudermanicus normally reproduces vegetatively by turions, it would also be interesting to know whether the small fruits produced by occasional flowers of the hybrid are ever capable of germination. Occasionally, Potamogeton hybrids may be found at sites where both parents are absent. P. acutifolius not only has a very restricted distribution in the Wareham area but is also much less frequent in those ditches where it does occur than it is at some sites in Sussex where it is present in abundance (Preston & Croft 1997). This suggests that we may be witnessing a late stage in the process by which a hybrid which has become established in the vicinity of its parents is then left alone following the disappearance of both parents from the locality. ACKNOWLEDGMENTS We are very grateful to Sir George Taylor for making his unpublished paper available to C.D.P. and for his permission to use this and other papers in our work on Potamogeton. A. O. Chater kindly extracted relevant passages from Salter’s diaries, Mrs O. M. Geddes and H. J. Noltie gave us information about the current whereabouts of Sir George Taylor’s papers, R. Vickery helped us with N. D. Simpson’s papers at BM and Dr T. C. G. Rich provided details of herbarium material at NMW. Dr H. J. M. Bowen, J. Fant, Lady Rosemary FitzGerald, M. Gurney, Miss A. Horsfall, Dr A. E. Newton and Mrs A. V. Pearman joined us to search for P. x sudermanicus at Wareham, and J. Day and B. Pickess provided us with information on plants north of the River Frome and access to the R.S.P.B. reserve. We thank Dr & Mrs L. A. Boorman for translating passages from Ploeg (1987) and D. B. Roy and Miss A. Stewart for generating Figs 1 and 2 respectively. REFERENCES AINSWORTH, G. C. (1986). John Ramsbottom (1885-1974), in BLAKE, LorD & NICHOLLS, C. S., eds. The dictionary of national biography 1971-1980, pp. 701-702. Oxford University Press, Oxford. BANCE, H. M. (1946). A comparative account of the structure of Potamogeton filiformis Pers. and P. pectinatus L. in relation to the identity of a supposed hybrid of these species. Transactions of the Botanical Society of Edinburgh 34: 361-367. CHARLTON, W. A. & PosLusZzNy, U. (1991). Meristic variation in Potamogeton flowers. Botanical journal of the Linnean Society 106: 265-293. ; DANDY, J. E. (1975). Potamogeton L., in STACE, C. A., ed. Hybridization and the flora of the British Isles, pp. 444—- 459. Academic Press, London. Danby, J. E. & TayLor, G. (1946). An account of x Potamogeton suecicus Richt. in Yorkshire and the Tweed. Transactions of the Botanical Society of Edinburgh 34: 348-360. Danby, J. E. & TayLor, G. (1957). Two new British hybrid pondweeds. Kew bulletin 1957: 332. FERNALD, M. L. (1932). The linear-leaved North American species of Potamogeton, Section Axillares. Memoirs of the American Academy of Arts and Sciences 17: 1-183. Goop, R. (1948). A geographical handbook of the Dorset flora. The Dorset Natural History and Archaeological Society, Dorchester. HaGstrOM, J. O. (1916). Critical researches on the Potamogetons. Kungliga svenska Vetenskapsakademiens Handlingar 55(5): 1-281. HOLLINGSworrtH, P. M., PRESTON, C. D. & GORNALL, R. J. (1995). Isozyme evidence for hybridization between Potamogeton natans and P. nodosus (Potamogetonaceae) in Britain. Botanical journal of the Linnean Society 117: 59-69. PLOEG, D. T. E. VAN DER (1987). Potamogeton x sudermanicus Hagstrém, de vermoedelijke bastaard van P. acutifolius en P. berchtoldii Fieber, in Nederland gevonden. Gorteria 13: 173-176. 172 C. D. PRESTON AND D. A. PEARMAN PRESTON, C. D. (1993). Irish pondweeds IV. Potamogeton x lanceolatus Smith. Irish naturalists’ journal 24: 213- 218. PRESTON, C. D. (1995a). Sir George Taylor’s studies of the genus Potamogeton. Watsonia 20: 325-326. PRESTON, C. D. (1995b). Pondweeds of Great Britain and Ireland. Botanical Society of the British Isles, London. PRESTON, C. D. & CRort, J. M. (1997). Aquatic plants in Britain and Ireland. Harley Books, Colchester. SELL, P. D. ed. (1967). Taxonomic and nomenclatural notes on the British flora. Watsonia 6: 292-318. STEARN, W. T. (1981). The Natural History Museum at South Kensington. Heinemann, London. (Accepted February 1998) Watsonia 22: 173-179 (1998) 173 Genetic conservation of Black poplar (Populus nigra L.) P. TABBUSH Forest Research, Alice Holt Lodge, Wrecclesham, Farnham, Surrey, GU10 4LH ABSTRACT Current interest in the conservation of Populus nigra L. (Black poplar) (Salicaceae) in Britain is set in a European context. A Black poplar network, set up under the European Forest Genetic Resources Programme, is described, and the reasons for choosing Black poplar as one of the first forest tree species to conserve are considered. The botanical status of Black poplar and its status in Britain are described, and against this background European and British strategies for genetic conservation of the species are discussed. Conclusions are drawn on the need to conserve the British native tree flora. KEYWORDS: taxonomy, characteristics, status, Britain. INTRODUCTION The continued health of managed forests (i.e. virtually all European forests) depends on the availability of suitable genetic material. Genetic diversity allows organisms to adapt to changes in the environment, and to changing patterns of predation and pest attack, so the irretrievable loss of genetic diversity in forest trees reduces our ability to maintain healthy, vigorous forests in the future. Until recently, mankind has taken genetic diversity for granted, largely because the scale of our activities was not sufficient to pose a serious threat. This is no longer the case; natural populations of forest trees are under threat on a global scale. In Britain, only about 1% of our forests are classified as “ancient and semi-natural” (Peterken 1981) with native species occupying sites which have been under forest since at least 1600 A.D. This is the first place to look for natural genetic diversity, because plantation forests are likely to contain trees which have been selected for timber producing characteristics, and which may well have been imported from elsewhere in the world. So, in Britain, the genetic diversity of our native trees is undoubtedly threatened. The European Forest Genetic Resources Programme (EUFORGEN) is a collaborative programme among European countries aimed at ensuring the effective conservation and sustainable utilization of forest genetic resources in Europe. It was established to implement Resolution 2 of the Strasbourg Minis- terial Conference on the Protection of Forests in Europe in 1990. The programme of EUFORGEN oper- ates through a small number of networks, and one of the first of these to be set up was the Populus nigra network (Frison et al. 1995). Fig. 1 illustrates a typical example of Black poplar. In setting about the genetic conservation of forest trees, it is perhaps surprising to find that Black poplar has been seen as a good place to start. The reasons for choosing Black poplar are: a. Its natural habitat — the floodplain forest — has been widely cleared to provide productive agricultural land; b. Control of rivers is widespread, and this disallows the conditions necessary for natural regeneration; c. It is easy to propagate by cuttings; d. It is one parent of the commercially important hybrid Populus x canadensis (syn. P. x euramericana (Dode) Guinier) (P. deltoides x P. nigra); e. Jt is highly resistant to bacterial canker (Xanthomonas populi) and shows some resistance to other important diseases of poplar; : f. Ona European scale, poplar is highly significant economically. For example, in France, about 35% of all hardwood timber produced is poplar, although the species only occupies about 1% of the area of broadleaved woodland; and g. It is relatively easy to identify which trees to conserve. 174 P. TABBUSH FiGuRE |. Mature Black poplar (Populus nigra) growing at Crickhowell in South Wales (May 1994). This tree recently blew down and has been removed. This last point requires some explanation. In the absence of clear information about the genetic make- up of a given population, and about the origins of these genes, it is difficult to be clear about whether the population is “native” or not. It is also difficult to define the boundaries of the block of genetic material which is to be conserved. The Common or Pedunculate oak (Quercus robur), for instance, presents a much more confused picture than Black poplar. Gardiner (1974), before DNA analysis was available, reviewed the evidence for and against the native status of the species, and of Sessile oak (Q. petraea). In his concluding remarks he indicated that Pedunculate oak has been planted in Scotland, and this has resulted in much introgression with the native Sessile oak, whilst in southern England it seems reasonable to suggest that both species are native. Using analysis of chloroplast DNA, Ferris et al. (1997) were able to distinguish an East Anglian population of Pedunculate oak from the many introduced specimens, and also found a marker which distinguished eastern and western European populations of both species of oak. This also showed that a number of very old oaks in Britain were in fact introductions from eastern Europe. The genetic diversity of our Common oak will therefore be difficult to conserve without first investigating its origins in detail using modern techniques of molecular genetics. BOTANICAL STATUS Populus nigra belongs to the family Salicaceae (poplars and willows). Within the genus Populus, it belongs to the section Aigeiros (Jobling 1990) along with P. deltoides, although study of the chemical composition of bud exudates has placed it in the section Tacamahaca (Greenway et al. 1990) along with P. trichocarpa. Varieties of P. nigra were listed by Zsuffa (1974), and the first two varieties distinguished (as recorded by that author) are: Variety: Synonyms: 1. P. nigra var. typica P. nigra L. (1753), P. nigra Dode (1905), P. europaea Dode Schneid. (1904) (1905), P. nigra var. Dodeana A. et G. (1908), P. nigra var. europaea A. et G. (1908) GENETIC CONSERVATION OF BLACK POPLAR 175 2. P. nigra var. betulifolia P. hudsonica Michx. (1813), P. betulifolia Pursh. (1814), (Pursh) Torr. (1843) P. pubescens Pursh (1814), P. nigra var. viridis (1838), P. nigra var. pubescens Parl. (1867), P. nigra var. hudsonica Schneid. (1904), P. Henryana Dode (1905), P. vistulensis Dode (1905), P. vaillantiana Dode (1905), P. Mulleriana Dode (1905), P. Lloydii Henry (1913). “P. nigra var. typica”’ is no longer used, “P. nigra L.” being used to describe the species as a whole. The accepted name for the second variety or subspecies listed above is now Populus nigra L. subsp. betulifolia (Pursh) Dippel. (Stace 1997), and this is the tree which is considered to be native in Britain. Throughout its world distribution (Fig. 2) P. nigra is typical of the alluvial forests of large European and Siberian rivers. According to Zsuffa (1974), subsp. betulifolia has been described in western Europe, especially in France and Great Britain, and “‘a pistillate form P. Lloydii Henry, is cultivated in England”’. Possession of the following characteristics distinguish Black poplar from its hybrids: Leaning trunk with large, swollen bosses; Deeply fissured dark bark; Absence of glands at leaf bases (of mature leaves); Presence of Pemphigus galls on the petioles; Complete resistance to mistletoe; Large, downward arching branches; Yellowish terete twigs, shining, becoming greyish; and Long reddish-brown buds curving outwards at the apex. Si ho ao os P. nigra subsp. betulifolia is distinguished by having hairs on the juvenile shoots, leaf petioles and midribs, and main flower stalk. The leaves are somewhat smaller (although leaf size varies on individual trees), and new growth is said to start earlier (Krussman et al. 1986). It remains to be seen if this subspecies can be clearly distinguished using DNA analysis. This will be important in the current context, as we will wish to know whether the British population (or the entire subspecies) should be conserved as something different and separate from P. nigra as a whole. STATUS OF BLACK POPLAR IN BRITAIN At about the time that the EUFORGEN Black Poplar Network was set up, concern for the future of P. nigra was growing in the UK (White 1993). It is no longer of importance in Britain as a timber tree Ves. | a9 a | ae on | 60 as, papi eS Ge z A , ee 45 30 fe sa “ads di — a : 15 : be y) FIGURE 2. Natural world distribution of Populus nigra L. (after Zsuffa 1974). 176 P. TABBUSH in its own right, and it is so rare that it is no longer a generally recognised feature of our landscape. It certainly has been an important feature of certain landscapes — John Constable’s famous painting ““The Hay Wain”’ painted in 1821 shows mature Black poplars, and in all probability Black poplar was used in the manufacture of the hay wain itself. Serious recording of the distribution of Black poplar started with the Botanical Society of the British Isles (B.S.B.I.) survey in 1973 (Milne-Redhead 1990). Milne- Redhead records that the Atlas of the British flora (Perring & Walters 1962) recorded a confusion of Black poplar and poplar hybrids, and the survey set out to clarify the true distribution of Black poplar, with the assumption that the old Black poplars (> 150 years old) were “native’’. The Flora of the British Isles (Clapham, Tutin & Warburg 1987) records Black poplar as “Native. Along river valleys south of the Mersey and Humber, but only introduced in Cornwall and W. Wales...”. According to Milne-Redhead (1990), all the current population derives entirely from planted stock. This is because the establishment of Black poplar from seed requires the seed to fall on bare wet mud, soil or silt at the end of June, and the site has to remain bare and wet until leaf-fall in October. Such conditions have vanished with the taming of rivers and the conversion of floodplain forests to agriculture. However, the most obvious source of cuttings would have been pre-existing native trees, and because the planted trees are very generally of subsp. betulifolia, it seems reasonable to assume that they are descendants from native trees. The pollen record gives us little information on this since the pollen of Black poplar is indistinguishable from that of Aspen (P. tremula L.) and White poplar (P. alba L.) (Huntley & Birks 1983). The database created by the B.S.B.I. catalogues over 2000 trees, and is still being maintained at the Biological Records Centre at the Institute of Terrestrial Ecology at Monks Wood. Records of Black poplar are verified and recorded by 6-figure Ordnance Survey grid reference and vice-county. Over 200 are positively identified as female, but the majority of trees are “‘unsexed”’. Rogers (1995) is therefore pessimistic in asserting that “there are only about 150 female trees in Britain’. The first hybrid between European P. nigra and the American P. deltoides introduced to Britain was probably ‘Serotina’, and this was introduced about 200 years ago (Jobling 1990). At the same time, P. nigra ‘Italica’, the familiar Lombardy poplar, was introduced from Northern Italy, although in both cases in would have taken some time for specimens to become widely distributed. Since both varieties are male, we cannot be certain that any seedlings dating from later than this are free from imported genes. Therefore, the database contains records of “‘mature”’ trees, which are difficult to date precisely, but which are probably more than 150 years old. ‘Manchester Poplar’ is a horticultural synonym for P. nigra var. betulifolia (Stace 1971; Bean 1976) but it has come to refer to urban trees planted in Manchester and the industrial north generally. These urban trees have been largely excluded, since Stace (1971) examined 100 specimens and found that they were all male, and therefore possibly representing a very narrow genetic base propagated vegetatively. However, since there is a great preponderance of males among the British population generally, this hardly seems to be adequate grounds for the exclusion of the urban trees, and they should be included in DNA studies of the British population. GENETIC STUDIES Modern molecular methods for analyzing genetic material now offer means for examining genetic diversity, and defining relationships between individual taxa. Legionnet (1997) used isozymes to examine genetic diversity and population biology in P. nigra growing in France, and found that there was more genetic diversity within rather than between stands. As a consequence, it would be more efficient to conserve more individuals from a small number of stands than vice versa. Cottrell et al. (1997) used RAPD markers to study 36 accessions of Black poplar broadly sampled within Great Britain, and found only 17 distinct genotypes. Genotypes were local in their distribution and genetic diversity was low. These authors also concluded that there had been so much interference by man that there are unlikely to be distinct Eastern and Western types. In a more concentrated study of Black poplar in the Upper Severn area, Winfield et al. (1998) used AFLP analysis to examine genetic diversity in 146 individuals and three individuals considered to be non-betulifolia poplars. Genetic diversity was low, confirming the results of Cottrell et al. (1997). There was a general correlation between geographic proximity and genetic similarity. They concluded that it was possible to identify a GENETIC CONSERVATION OF BLACK POPLAR 177 small number of individuals exhibiting maximum diversity for inclusion in a replanting/conservation programme. Of the 36 trees sampled by Cottrell et al. (1997), only six were female, and DNA analysis of these revealed that there were only two distinct genotypes, despite the fact that they were sampled from a wide geographical range. METHODS OF GENETIC CONSERVATION The Biodiversity Action Plan (Anon. 1994), was written following the UK signature to the Convention on biodiversity at the United Nations Conference on Environment and Development at Rio de Janeiro in 1992. It distinguishes between in situ and ex situ conservation measures, and states that ‘“‘for successful conservation the establishment of ex situ populations should precede any crisis period and should be implemented when wild populations are still quite numerous.” In the case of Black poplar, we can assume that the in situ population (in Britain) is entirely of planted origin. We might seek to conserve standing trees because they have not yet been identified and conserved, or because they have particular cultural associations, e.g. the Aston-on-Clun Flag Tree (Mabey 1996), but the principal work must be in identifying the genetic diversity of the taxon (species or subspecies) and its distribution, and seeking the most efficient way to conserve it in ex situ collections. There is not much time available for this process, since there is significant loss of the oldest and most valuable trees, at a rate which has not yet been measured. It could be measured by revisiting a sample of trees taken from the B.S.B.I. database at regular intervals. Meanwhile, a clone bank has been built up by the Forestry Commission, but this contains less than 100 accessions and, on the basis of the molecular studies, might be expected to contain less than half this number of distinct genotypes. No collection has yet been made based on genetic information of the type obtained for the River Severn area (Winfield et al. in press). The current collection is held by the Forestry Commission at three separate sites-(for security) in Bedfordshire, Norfolk and Gwent. In all cases, the trees are planted at wide spacing, with the intention of allowing them to grow into mature trees. Another strategy is to hold reference collections in closely- spaced stool-beds, which are cut annually, so that it is relatively easy to provide cuttings, e.g. for DNA studies. Part of the National Poplar Collection at Alice Holt (Surrey) is already managed in this way, and it would be relatively easy to add a Black poplar collection. Such a collection would be more efficient if we could be certain that it contained no duplicate genotypes. This could be achieved following a broad survey of genetic variability (using DNA analysis) in putatively native standing trees. EUROPEAN CONSERVATION STRATEGY The P. nigra network of EUFORGEN has adopted a workplan (Turok et al. 1996) with the following activities: 1. Exchange of reference clones: a common set of vegetative material of well-known poplar clones has been distributed to interested countries for use as a base-line in genetic studies. 2. Core collection of P. nigra: vegetative material from identified clones in member countries has been collated at the Poplar Research Institute at Casale Monferrato in Italy. The aim is to set up a common basis for the characterization and evaluation of national clone collections. 3. “Passport data”: an agreed list of attributes has been compiled to identify material in a database, and for inventories and exchange of material. 4. European database: an organised list of accessions available in each country has been compiled to avoid confusions about clonal identity. 5. Descriptor list for P. nigra clones: the aim is to produce common minimum standards for characterizing clone collections in each country. 6. Identification sheet: an illustrated sheet (Turok et al. 1996) is available as a simple guide to distinguish P. nigra from its hybrids (though it does not go down to subspecies level). Synthesis of in situ gene conservation measures and activities: published in Turok ef al. (1996). Guidelines for ex situ field collections: published in Turok ef al. (1996). Cora 178 P. TABBUSH 9. Guidelines for seed and pollen storage: published in Turok et al. (1997). 10. Review of literature: initial version in Frison et al. (1994). Updated versions to be made available through the Internet. 11. Public awareness: assembly of colour slides related to P. nigra. 12. Country reports: published in the various reports, and updated regularly. 13. Molecular methods available for the characterization of P. nigra: to review and develop modern methods for the characterization of existing collections. Not yet published. This list indicates the breadth of activity that has been deemed necessary to conserve Black poplar. Similar progress is being made by other EUFORGEN networks on Picea abies, Quercus suber and the ‘Noble Hardwoods”. This last category includes alder, ash, Sorbus, elm and fruit woods, but excludes oak and beech, which are classified as “Social Hardwoods” presumably because they tend to form large single-species stands. U.K. NATIONAL STRATEGY FOR THE GENETIC CONSERVATION OF TREE SPECIES Efforts to prevent the disappearance of our most endangered species of timber tree have highlighted the need to ensure that the genetic base of all our native trees is not eroded. Britain has a native tree flora of only 33 species (Mitchell 1981). The natural distributions of these species and their genetic diversity are still unknown or imperfectly understood. Indeed, in many cases, it is difficult to tell which trees are “native” and which have been imported by man. A strategy for conserving the genetic diversity of our native trees must first organise the existing information, and then seek to identify which individuals should be protected. A start on this has already been made in relation to Black poplar, and in collaboration with the EUFORGEN P. nigra network. REFERENCES ANON. (1994). Biodiversity: the UK action plan. H.M.S.O., London. BEAN, W. J. (1976). Trees and shrubs hardy in the British Isles, 8th ed. John Murray, London. CLAPHAM, A. R., TUTIN, T. G. & WARBURG, E. F. (1987). Flora of the British Isles, 2nd ed. Cambridge University Press, Cambridge. COTTRELL, J. E., FORREST, G.I. & WuiTeE, I. M.S. (1997). The use of RAPD analysis to study diversity in British black poplar (Populus nigra L. subsp. betulifolia (Pursh.) W. Wettst. (Salicaceae)) in Great Britain. Watsonia 21: 305-312. FERRIS, C., Davy, A. J. & Hewitt, G. M. (1997). A strategy for identifying introduced provenances and translocations. Forestry 70, 211-222. FRISON, E., LEFEVRE, F., DE VRIES, S. & TUROK, J. compilers (1995). Populus nigra Network. Report of first meeting, 3—5 October 1994, Izmit, Turkey. IPGRI, Rome, Italy. GARDINER, A. S. (1974). A history of the taxonomy and distribution of the native oak species, in Morris, M. G. & PERRING, F. H. eds, The British oak. B.S.B.I., Classey, London. GREENAWAY, W., ENGLISH, S. & WHATLEY, F. R. (1990). Variation in bud exudate composition of Populus nigra assessed by gas chromatography-mass spectrometry. Zeitschrift fiir Naturforschung 45, 931-936. HunTLEY, B. & Birks, H. J. B. (1983). An atlas of past and present pollen maps for Europe: 0-13,000 years ago. Cambridge University Press, Cambridge. JOBLING, J. (1990). Poplars for wood production and amenity. Forestry Commission bulletin 92. H.M.S.O., London. K RUSSMAN, G., Epp, M. & DANIELS, G. (1986). Manual of cultivated broadleaved trees and shrubs. Batsford, London. LEGIONNET, A. (1997). in TUROK, J., LEFEVRE, F., DE VRIES, S. & ToTH, B. Populus nigra Network. Report of third meeting, 5—7 October 1996, Sarvar, Hungary. IPGRI, Rome, Italy. MAaABEY, R. (1996). Flora Britannica. Sinclair-Stevenson, London. MILNE-REDHEAD, E. (1990). The B.S.B.I. black poplar survey, 1973-88. Watsonia 18: 1-5. MITCHELL, A. F. (1981). The native and exotic trees in Britain. Arboricultural Research Note 29/81/SILS. PERRING, F. H. & WALTERS, S. M., eds. (1962). Atlas of the British Flora. Thomas Nelson & Sons, London. PETERKEN, G. F. (1981). Woodland conservation and management. Chapman and Hall, London. Rocers, E. (1995). The native black poplar (Populus nigra var. betulifolia (Pursh.) Torr.) Quarterly journal of forestry 89: 33-37. GENETIC CONSERVATION OF BLACK POPLAR 179 STACE, C. A. (1971). The Manchester poplar. Watsonia 8: 391-393. STACE, C. A. (1997). New Flora of the British Isles, 2nd ed. Cambridge University Press, Cambridge. TuROK, J., LEFEVRE, F., CAGELLI, L. & DE VRIES, S. compilers (1996). Populus nigra Network. Report of the second meeting, 10-12 September 1995, Casale Monferrato, Italy. IPGRI, Rome, Italy. TuROK, J., LEFEVRE, F., DE VRIES, S. & ToTH, B. compilers (1997). Populus nigra Network. Report of the third meeting, 5—7 October 1996, Sarvar, Hungary. IPGRI, Rome, Italy. Waite, J. (1993). Black Poplar: The most endangered native timber tree in Britain. Forestry Commission research information note 239. WINFIELD, M. O., ARNOLD, G. M., Cooper, F., LE Ray, M., WHITE, J., KARP, A. & EDWARDS, K. J. (1998). A study of genetic diversity in Populus nigra subsp. betulifolia in the upper Severn area of the U.K. using AFLP markers. Molecular ecology 7: (in press). ZsuFFA, L. (1974). The genetics of Populus nigra L. Annales Forestales Anali Za Sumartvo 6/2. (Accepted February 1998) ae < ak! ey Te | sy pty % ie Ki i i , ee es a Ay % chepeat a Watsonia 22: 181-193 (1998) 181 Notes TYPIFICATION OF SOME BARTON & RIDDELSDELL NAMES IN RUBUS L. (ROSACEAE) In the two successive papers in which they described Rubus bakerianus, R. furnarius, R. newbridgensis and R. pistoris Barton & Riddelsdell (1935, 1936) indicated holotypes identified by numbers in Barton’s herbarium. The gatherings to which those numbers referred were respectively: Wimbledon Common, v.c. 17, 3 August 1934, Barton & Riddelsdell 4376 Sedbergh, v.c. 65, 1934, Barton & Riddelsdell 4378 Newbridge, v.c. 14, 2 August 1934, Barton & Riddelsdell 4519 Gormire, near Thirsk, v.c. 62, 1 August 1933, Riddelsdell 4351. Unfortunately, the two were inadequately versed in the typification procedure prescribed in the new edition of the /nternational rules (Briquet 1935) which had just then come into force. This explicitly stated that a holotype must consist of “a specimen” (Art. 18, Rec. IV). A specimen for the purposes of the microspecies of Rubus fruticosus agg. has long been accepted by specialists in this group as necessarily consisting of a flowering spray together with part of the primocane with at least one stem leaf attached to that. It is conventionally regarded as unsafe to venture a determination in the absence of either of those components. While Barton & Riddelsdell observed this convention impeccably, after Barton’s herbarium (incorporating Riddelsdell’s) passed to BM after his death it emerged that the designated holotypes in each of these four cases consisted not of a single specimen but of several. In the case of R.. bakerianus as many as five sheets of specimens turned out to be sharing the particular number cited, three of those sheets bearing more than one flowering spray a piece. That in each case the specimens comprising the so-called holotype all came from the same bush is rendered likely by the fact that Barton labelled some of the sheets “co-type’’ (a term without official standing in the /nternational rules either then or since); however, that word is not present on all, leaving open the possibility that more than the one bush may have been involved. As more precise typification was clearly called for, at some unstated date one appropriate sheet was chosen in each case as the lectotype and labelled as such in an anonymous hand. Unfortunately, the fact that this had been done was not published, allowing Edees & Newton (1988), in their monograph of the group in the British Isles, to repeat from Barton & Riddelsdell’s papers the details of what they supposed to be particular single specimens and similarly to cite these as holotypes. The handwriting on the labels has now been identified as that of G. A. Matthews, a former member of the BM British Herbarium, and to him the lectotypifications are now here belatedly credited — except in the case of R. pistoris. Of that species the sheet in question bears three flowering spray pieces and five primocane pieces, making it necessary to restrict the lectotypification further to just one pair of those. The lower right-hand spray plus associated primocane piece (as indicated on the sheet) are accordingly here so designated. ACKNOWLEDGMENTS I am grateful to Dr C. E. Jarvis for advice on this matter and to A. R. Vickery for identifying the handwriting in question. REFERENCES BARTON, W. C. & RIDDELSDELL, H. J. (1935). Rubus bakeri F. A. Lees and its allies. Journal of botany 73: 124-131. BARTON, W. C. & RIDDELSDELL, H. J. (1936). A new Rubus from Sussex. Journal of botany 74: 204-205. BRIQUET, J., ed. (1935). International rules of botanical nomenclature, 3rd ed. Fischer, Jena. EDEES, E. S. & NEwTON, A. (1988). Brambles of the British Isles. Ray Society, London. D. E. ALLEN Lesney Cottage, Middle Road, Winchester, Hampshire, SO22 5EJ 182 NOTES TRIFOLIUM OCCIDENTALE D. E. COOMBE (FABACEAE) IN ANGLESEY (V.C. 52) The Western Clover, Trifolium occidentale D. E. Coombe, is an early-flowering and self-compatible coastal species, which is stoloniferous and superficially resembles T. repens L. Since it was first described by Coombe (1961), detailed information on the narrow maritime Atlantic range of T. occidentale has been reported at intervals, as discoveries of new localities have accumulated. In a recent summary review with accompanying map, Coombe (1994) outlined its specialised ecological distribution in exposed coastal-fringe grasslands in the Channel Isles, the Isles of Scilly and mainland W. Cornwall, with isolated records in N. Devon and on the Gower peninsula in Glamorgan; he was also able to report 7. occidentale as frequent in south-west Ireland and north-west France, and rare on the west Iberian coast. It has since been recorded by R. S. Cropper at three localities in south-west Pembrokeshire (Evans 1997). A new record from Anglesey outlined in this note marks a further northward extension of its range on the west coast of Britain. The most northerly locality for T. occidentale in Europe is in Co. Dublin where it was reported by Preston (1980) and Akeroyd (1983). In May 1995, Alan Lewis informed RHR (then vice-county recorder for v.c. 52) that he had recently found a few plants which resembled T. occidentale near Trearddur Bay, on Holy Island, in west Anglesey. A subsequent visit to the locality was delayed until April 1997, when we were able to examine plants and confirm them as undoubted T. occidentale. Very shortly afterwards, the site was visited by another party of botanists including C. D. Preston who also confirmed this determination. The Anglesey population of T. occidentale is evidently quite small and highly restricted in extent. On 19 April 1997, about 20 separate patches were observed on a low bank above a small car-park at the landward end of a short rocky headland forming the south side of Porth Diana (SH/253.782). The soil of the bank has a high fraction of wind-blown sand; it is somewhat unstable and was probably disturbed during construction of the car-park. A single plant was also noted by a gateway on the opposite side of the minor road from the car-park. There is no evidence to indicate whether T. occidentale is a long- established native or if it has recently arrived in west Anglesey. The population is close to the rocky shore-line and, as in other parts of its range, some of the associated species at Porth Diana are strongly maritime. Among the species growing in close proximity to T. occidentale are Anthyllis vulneraria L., Bellis perennis L., Bromus hordeaceus L., Carex arenaria L., Catapodium marinum (L.) C. E. Hubb., Cerastium diffusum Pers., Cochlearia danica L., Erodium cicutarium (L.) L’Hér., Festuca rubra agg., Galium verum L., Hypochaeris radicata L., Lotus corniculatus L., Medicago lupulina L., Ononis repens L., Plantago coronopus L., P. lanceolata L., Poa pratensis agg., Ranunculus bulbosus L., Sanguisorba minor L., Scilla verna Huds., Senecio jacobaea L. and S. vulgaris L. . TABLE 1. COMPARISON OF SEED SET AND SEED SIZE IN TRIFOLIUM OCCIDENTALE AND T. REPENS FROM ANGLESEY, v.c. 52 a. Seed set No. of seeds per flower* 0 ] 2 3 4 T. occidentale (29 June 1997) 55 Sil 30 6 ~ T. repens (5 August 1997) 51 24 14 10 b. Seed size Seed length (mm)** Mean (range) n T. occidentale 1-2 (1-:0—1-6) Lv T. repens 1-1 (0-8-1-4) 88 *Counts of seed set for samples of 100 flowers in each species, obtained from ten inflorescences, each from a separate individual. **Seed length data obtained from samples of ten inflorescences per species, each from a separate individual. FiGure |. The distribution of Welsh records of Trifolium occidentale.@ Occurrence in 5 x 5 km squares of the national grid. When it was observed on 19 April 1997, a few plants of T. occidentale had early inflorescences with flowers in bud. The locality was visited on several subsequent occasions during 1997 to make comparisons between T. occidentale and the much larger population of T. repens present in coastal turf on the headland. The main flowering period of T. occidentale lasted until mid-June, with a few stragglers appearing beyond this period. In contrast, T. repens started to come into flower in late May and early June, with its main flowering period persisting through to early August. Material gathered from several plants of 7. occidentale showed little variation. They agreed with | descriptions given by Coombe (1961) and Coombe & Morisset (1967), having small (c. 10 x 10 mm), orbicular leaflets, of a darker green than in 7. repens and without the leaf-markings common in that species; their upper leaflet surface is matt and minutely crystalline when viewed under low-power magnification, and the underside is strikingly glossy; the lateral veins are characteristically non- 184 NOTES translucent when a leaflet is viewed against strong diffuse light. In addition, the stipules are a deep vinous red, which was particularly apparent in the young creeping shoots. This set of vegetative characteristics is diagnostic of T. occidentale. The presence of sparse, short, colourless hairs on the petioles and peduncles has often been cited as an additional distinctive character, but in our experience they also occur frequently in T. repens. Floral characters in the Anglesey plants were also found to be equally uniform and in full agreement with those described by Coombe (1961). The flowers are creamy white with no tinge of pink, the standard is broadly elliptical and emarginate (a very shallow emargination also occasionally occurs in T. repens); the upper teeth of the calyx are parallel or convergent, broadly triangular or ovate-triangular and often minutely denticulate. As seed pods ripen, the flowers become a dark chocolate brown in T. occidentale with no trace of pink or red pigment which frequently suffuses the fruiting heads of T. repens. Comparative estimates of seed set and seed size were made between T. occidentale and T. repens. Seed counts were recorded from samples of ten flowers taken from each of ten inflorescences per species from different plants; data were obtained in the main fruiting period in each species, and seeds damaged by insect attack or by pathogens were ignored. For both species seed length measurements were made on samples from ten inflorescences, each from a separate plant. The results are given in Table 1. Seed germination was tested in two plants of T. occidentale, sown on 30 June 1997. Of 16 seeds sown from Plant A, 15 germinated between 13-24 July, while four seeds out of 13 sown from Plant B germinated between 5—10 August. These findings suggest that seed size and productivity are similar in the two species, and also that at least a proportion of the T. occidentale seed is viable. We have searched several other headlands and stretches of rocky coast in south-west Anglesey, extending from Holy Island to Llanddwyn Island, but have found no further colonies of T. occidentale. It also appears that it is very scarce on the Gower peninsula (Kay & Ab-Shukor 1988) and elsewhere in Wales; the only known records are shown in Fig 1. ACKNOWLEDGMENTS We thank Alan Lewis for passing on his observations in 1995, Gwynn Ellis for providing information about Welsh records, and Chris Preston for his observations. REFERENCES AKEROYD, J. R. (1983). Further notes on Trifolium occidentale D. E. Coombe in Ireland. /rish naturalists’ journal 21: 32-34. CoomekE, D. E. (1961). Trifolium occidentale, a new species related to T. repens L. Watsonia 5: 68-87. CoomBE, D. E. (1994). Trifolium occidentale Coombe, in STEWART, A., PEARMAN, D. A. & PRESTON, C. D. eds. Scarce plants in Britain, p. 414. Joint Nature Conservation Committee, Peterborough. CoomBE, D. E. & Morisset, P. (1967). Further observations on Trifolium occidentale. Watsonia 6: 271-275. EvANS, S. B. (1997). Welsh plant records — 1995. Pembroke v.c. 45. Botanical Society of the British Isles, Welsh bulletin 62: 33-34. Kay, Q. O. N. & AB-SHuKOR, N. A. (1988). Trifolium occidentale D. E. Coombe, new to Wales. Watsonia 17: 168-170. PRESTON, C. D. (1980). Trifolium occidentale D. E. Coombe, new to Ireland. /rish naturalists’ journal 20: 37-40. T. H. BLACKSTOCK Countryside Council for Wales, Penrhos Road, Bangor, Gwynedd, LL57 2LQ R. H. ROBERTS 51 Belmont Road, Bangor, Gwynedd, LL57 2HY LECTOTYPIFICATION OF POTAMOGETON FLABELLATUS BAB. (POTAMOGETONACEAE) Since it was first described by Babington (1851), the name Potamogeton flabellatus has been applied to a variety of taxa. In Britain these have included broad-leaved plants which are now subsumed within NOTES 185 the variable P. pectinatus L. and superficially similar plants which are now known to be P. x suecicus K. Richt. (Dandy & Taylor 1946). Although the name is now usually reduced to synonymy, it survives as a convenient informal label for certain variants of P. pectinatus (Preston 1995). Lectotypification of the name is desirable both in itself and as part of the general need to typify names in the genus (Wiegleb 1988). C. C. Babington was one of the first British botanists to take a critical interest in the genus Potamogeton. In his Manual of British botany (1843) he published a serviceable account of the genus as it was then known. He split the species into five groups, one of which corresponds with the current Subgenus Coleogeton Rchb. In this group he recognised P. filiformis (which had not previously been reported from the British Isles) and he divided the plants which we would now call P. pectinatus into two taxa, P. pectinatus and P. zosteraceus Fr. He described P. zosteraceus as having linear-acuminate, obscurely 3-veined leaves and fruits rounded on the back with a prominent keel, whereas P. pectinatus had narrower, linear-setaceous, |-veined leaves and fruits with lateral ridges but no keel. He reported P. zosteraceus from only one site, the Serpentine in Hyde Park, London, where it had been collected by Dr J. A. Power. In the second edition of the Manual Babington (1847) retained P. zosteraceus, again on the basis of Power’s plant from the Serpentine. Although there are minor changes in wording, his description of the plant is essentially the same as in the first edition. He had, however, begun to doubt whether the plant he described was the same as that described by Fries (1828). In an additional note at the end of this group of species he also drew attention to plants from Bath and Somerset which were ‘probably a new species but I am not sufficiently acquainted with it to describe it’. By the time that he prepared the third edition of the Manual, Babington (1851) had concluded that Fries’ P. zosteraceus was not the plant that he had previously described under that name. He therefore introduced the name P. flabellatus for the British plant, with the synonym “P. zosteraceus Bab. (not Fr.)”. He rewrote the description of the plant, characterising it as having broadly linear, abruptly apiculate or acuminate 5-nerved lower leaves which are normally decayed at the time of flowering, narrow, acute 3-nerved upper leaves borne on stems which are branched and spread like a fan, and keeled fruits. The flowering period is given as June-July, whereas previously P. zosteraceus had been described as flowering in July. This description is more detailed than that of the earlier descriptions of P. zosteraceus and it places more emphasis on vegetative characters. In particular, the description of the habit and broad lower leaves is new. P. flabellatus was said to occur in “ponds and ditches”’ in England, but Babington cited no specific localities. The plants from Bath and Sandwich mentioned as a possible new species in the second edition are not explicitly cited. Although the concise entry for P. flabellatus in Babington’s Manual provides no indication of the source of the material he described, Babington (1853) later provided a more detailed account of his species. In this he stated that he described P. zosteraceus in the first two editions of the Manual on the basis of a plant ‘““which was very slightly known to me, it having been noticed in Hyde Park only”’. In 1849, however, he obtained ‘‘a series of most characteristic specimens” from Mr [Thomas] Kirk of Coventry, which convinced him that the plant was distinct from both P. pectinatus and from the true P. zosteraceus. “Accordingly, in the ‘Manual’ (ed. 3) the name of zosteraceus is changed into flabellatus, a term derived from the usually fan-shaped habit of the flowering plant.” At the time he prepared the account of P. flabellatus for the Manual he thought that the plant from Bath was referable to P. pectinatus, but visits to Bath in 1853 had enabled him to re-examine the plant and convinced him that it was in fact P. flabellatus. Babington makes no mention in this paper of the plant from Sandwich. It is clear from this historical resumé that the name Potamogeton flabellatus Bab. must be treated as a species described afresh in 1851, rather than simply as a replacement name for the plants from the Serpentine hitherto treated as P. zosteraceus. It is therefore appropriate to consider as syntypes all the material which Babington had available to him in 1851 and which he then considered referable to P. flabellatus. The following specimens from Babington’s herbarium (now incorporated into CGE) were collected before 1851 and labelled as P. flabellatus by Babington and are available for selection as the lectotype. The names on the specimens are given in the order in which they were applied; names in inverted commas were given by the collector and the rest are in Babington’s handwriting. A. Potamogeton pectinatus/zosteraceus/zosteraceus Bab. not Fries/flabellatus. Serpentine, Hyde Park, London. J. A. Power, 9 July 1838. B. “Potamogeton zosteraceus Bab.”’/flabellatus. Canal, Stoke Heath, Warwick. T. Kirk, June 1847. C. “Potamogeton zosteraceus Bab.” /flabellatus. Canal, Stoke Heath, Warwick. T. Kirk, 26 May 1849. 186 NOTES D. “Potamogeton zosteraceus Bab.”’/flabellatus. Canal, Stoke Heath, Warwick. T. Kirk, July 1849. E. Potamogeton zosteraceus/flabellatus. Coventry. T. Kirk, 29 May 1850. F. Potamogeton zosteraceus Bab.? River Lea below Ware, Herts. W. H. Coleman, 1848. Specimen A is the plant described under the name P. zosteraceus in the first two editions of the Manual. Specimens B-E represent the series of specimens Babington received from Kirk which convinced him that the plant was a distinct but undescribed species. “Warwick” on Kirk’s labels indicates the vice-county of Warwickshire; Stoke Heath is actually in Coventry and the canal which runs through it is the Coventry Canal. As specimen F was initially labelled “Potamogeton zosteraceus Bab.?”’, Babington must have received it before he coined the name P. flabellatus. In including it in the above list I have assumed that Babington decided that the specimen was P. flabellatus before he published this name, but I cannot prove it. Other than these six specimens, the only sheet in CGE collected before 1851 and labelled as P. flabellatus by Babington is a sheet from the “canal at Bath” collected by Babington himself in June 1830. This is the material which he referred to P. pectinatus at the time he described P. flabellatus (Babington 1853), and P. pectinatus is one of three names Babington wrote on the sheet and subsequently crossed out. It cannot, therefore, be regarded as a potential lectotype. An ideal lectotype of P. flabellatus would show both the fruits and the lower leaves, the two characters which Babington (1851) emphasised when describing the species. Unfortunately none of the syntypes show both these features, presumably because the lower leaves have usually decayed by the time that the plants flower (as Babington noted). Specimens B and F are upper flowering stems with neither lower leaves nor fruits, and need not be considered further. The choice therefore lies between the single specimen from the Serpentine (A) and the Stoke Heath specimens (C—E). The plant from the Serpentine (A) is cited indirectly in the protologue via the reference to Babington’s earlier description of P. zosteraceus. It was J. E. Dandy & G. Taylor’s choice of lectotype: it is cited as such in their unpublished monograph “The British species of Potamogeton L.”’, and they labelled the specimen in CGE “Type specimen of Potamogeton flabellatus Bab., Man. Brit. Bot. ed. 3, 343 (1851)”. This choice was never published, however, and does not constitute effective lectotypification (see International code of botanical nomenclature 1994, Article 7.10). Although the argument for selecting this specimen as lectotype is strong, I do not believe that it is conclusive. The protologue of P. flabellatus incorporates features which could not be derived from the Serpentine specimen, and it is clear that the Stoke Heath plants provided the main source for Babington’s revised description. It is appropriate, therefore, to select specimen C, D or E as the lectotype. It is better to select a fruiting plant rather than a vegetative shoot and I have therefore selected specimen C; it is an upper fruiting stem and the specimen includes one fruit which Babington removed and sectioned. Specimen D, which consists of broad lower leaves, is mounted on the same sheet. Specimen E, which is labelled “Spring leaves” consists of very broad lower leaves. All the syntypes were determined as P. pectinatus L. by J. E. Dandy and G. Taylor in 1939, determinations with which I concur. The nomenclature and typification of P. flabellatus can be summarised as: Potamogeton flabellatus Bab., Man. Brit. bot. ed. 3, 343 (1851). Lectotype: Canal, Stoke Heath, Warwick. T. Kirk, 26 May 1849, CGE (designated here). REFERENCES BABINGTON, C. C. (1843). Manual of British botany. John van Voorst, London. BABINGTON, C. C. (1847). Manual of British botany, 2nd ed. John van Voorst, London. BABINGTON, C. C. (1851). Manual of British botany, 3rd ed. John van Voorst, London. BABINGTON, C. C. (1853). On Potamogeton flabellatus Bab. Phytologist 4: 1158-1160. Danpy, J. E. & TAYLor, G. (1946). An account of x Potamogeton suecicus Richt. in Yorkshire and the Tweed. Transactions of the Botanical Society of Edinburgh 34: 348-360. FRIES, E. (1828). Novitiae Florae Suecicae, 2nd ed. Berling, Lund. PRESTON, C. D. (1995). Pondweeds of Great Britain and Ireland. Botanical Society of the British Isles, London. WIEGLEB, G. (1988). Notes on pondweeds — outlines for a monographical treatment of the genus Potamogeton L. Repertorium novarum Specierum Regni vegetabilis 99: 249-266. C. D. PRESTON I.T.E., Monks Wood, Abbots Ripton, Huntingdon, Cambs., PEI7 2LS NOTES 187 AQUATIC PLANTS AT HIGH ALTITUDES IN THE BREADALBANE MOUNTAINS (V.C. 88), SCOTLAND The standard work on the altitudinal range of British plants is Wilson (1956), which was based on papers originally published in 1930 and 1931 and is now somewhat dated. Some groups treated by Wilson have subsequently been the subject of taxonomic revision, and the altitudinal ranges of the taxa now recognised need to be established. There is also a need to localise many of the records cited. We believe that altitudinal limits should be based on precisely localised records, but in many cases Wilson (1956) relied on earlier authors who occasionally cited upper altitudinal ranges from areas as large and vague as the Scottish Highlands. Finally, there is a need to collate records made since Wilson’s compilation was published. In compiling data on the maximum altitude of aquatic plants in Britain and Ireland for publication (Preston & Croft 1997), it became clear that several species reached their upper limit in the Breadalbane mountain range. Many of these altitudinal limits are derived from White (1898). White usually stated the upper limits of species in Perthshire without citing the exact localities where these limits were reached, although for less common species these localities can often be deduced from the list of records. Our enquiries suggested that there were surprisingly few recent localised records of aquatic plants from these much-visited mountains: presumably botanists visiting the area have been preoccupied by the rich terrestrial montane flora and have seen no need to record aquatic plants which can be seen much more easily elsewhere. We therefore visited some lochs and lochans at high altitudes in the Breadalbane area in July 1995. The significant altitudinal records we made are detailed in this note. All the sites we visited in this area are in v.c. 88. We also refer to records made on a visit to Loch Vrotachan, v.c. 92, in 1996. In the records which follow, altitudes cited by earlier authors have been converted from feet to metres, and all altitudes are rounded to the nearest 5 m. The term lowland is used here for altitudes below 300 m; Wilson (1956) used it similarly, for altitudes below 1000 feet (305 m). The card index of Potamogeton specimens compiled by J. E. Dandy and held in BM is referred to as the “Dandy index’’. Unless stated the records quoted here from the index refer to specimens which Dandy cited from BM but which are not now incorporated into the herbarium, probably because they were lost or damaged in the Second World War (cf. Preston 1988). Nomenclature follows Stace (1991) for vascular plants and Moore (1986) for charophytes. Carex lasiocarpa. With C. rostrata in swamp around lochan N. of Lochan Achlarich, E. of Beinn Heasgarnich, altitude 650 m, NN/432.381, 26 July 1995. Although C. lasiocarpa often fails to flower or flowers very sparingly, this was not the case at this site in the summer of 1995, where flowers were frequent. Wilson (1956) follows White (1898) in giving the upper limit of this species as 425 m in the Atholl region of Perthshire. Carex limosa. Bog pools in flat-bottomed valley N.E. of Lochan Achlarich, E. of Beinn Heasgarnich, altitude 650 m, NN/436.385, 26 July 1995. Although this just exceeds White’s (1898) and Wilson’s (1956) upper limit, 640 m in Breadalbane, the species has been recorded by R. W. David at 830 m on Meall nan Tarmachan, NN/589.390, where it grew with C. saxatilis in a mire on a level shelf. This altitude, originally recorded as 2725 feet, is erroneously cited as 817 m by Jermy, Chater & David (1982); details of the original record are held at the Biological Records Centre. Carex nigra. Edge of lochan fed by two melting snowpatches, 200 m S.E. of the summit of Beinn Heasgarnich, altitude 1005 m, NN/415.382, 26 July 1995. Also present at lochans at 995 m (NN/ 417.384) and 970 m (NN/418.385) elsewhere in this area. The upper limit for this species cited by Wilson (1956) is based on Macvicar’s (1894) report of plants at 990 m within a radius of 10 miles [16 km] from Killin, v.c. 88. Carex rostrata. Swamp on flat ground between Meall Garbh and Beinn nan Eachan, altitude 930 m, NN/572.386, 25 July 1995. This exceeds Wilson’s (1956) upper limit for this species, 915 m, based on White’s (1898) unlocalised record from the Breadalbane area. However, D. A. Ratcliffe recorded this species at 1040 m E. of the main plateau of Creag Meagaidh, NN/432.875, on 2 August 1957 (cf. McVean & Ratcliffe 1962, pp. 116-117). Equisetum arvense. In Sphagnum at the edge of lochan fed by two melting snowpatches, 200 m S.E. of the summit of Beinn Heasgarnich, altitude 1005 m, NN/415.382, 26 July 1995, CGE (Preston 95/ 55). Also present by lochans at 995 m (NN/417.384) and 970 m (NN/418.385) elsewhere in this 188 NOTES area. The upper limit for this species given by Wilson (1956) is 945 m, based on Macvicar’s (1894) record from a radius of 10 miles [16 km] from Killin. Macvicar reported his plant as var. alpestre, a variant with short prostrate stems and suberect branches. The plants we collected would be covered by Babington’s (1881) description of this variant. Equisetum palustre. By large lochan S.W. of the summit of Meall nan Tarmachan, altitude 945 m, NN/ 581.387, 25 July 1995. This exceeds the unlocalised upper record of this species, 915 m in the Breadalbane area (White 1898; Wilson 1956); the species is also reported by McVean & Ratcliffe (1962, p. 351) from 915 m at NN/635.409 on Ben Lawers. Menyanthes trifoliata. Tiny plants at edge of lochan fed by two melting snowpatches, 200 m S.E. of the summit of Beinn Heasgarnich, altitude 1005 m, NN/415.382, 26 July 1995, CGE (Preston 95/54). Also present by a lochan 500 m N. of the summit, NN/418.385, at an altitude of 970 m. Swamp on flat ground between Meall Garbh and Beinn nan Eachan, altitude 930 m, NN/572.386, 25 July 1995. Wilson (1956), citing White (1898), gives 925 m in Breadalbane as the upper limit for this species. Myriophyllum alterniflorum. In water 20-30 cm deep at edge of Lochan an Tairbh-uisge, Meall nan Tarmachan, altitude 780 m, NN/591.396, 25 July 1995, CGE (Preston 95/52). Frequent in water c. 1 m deep over sand, E. end of Loch Vrotachan, v.c. 92, altitude 750 m, NO/124.784, 27 July 1996, CGE (Preston 96/96). Lochan Coire Dhubhclair, altitude 735 m, NN/497.329, 27 July 1995. These records exceed the upper altitudinal limit given by Wilson (1956), 715 m in Breadalbane and in Wales. Potamogeton alpinus. Water 16—50 cm deep over an otherwise bare schistose substrate, large lochan S.W. of the summit of Meall nan Tarmachan, altitude 945 m, NN/581.387, 25 July 1995, CGE (Preston 95/53). A large population of P. alpinus was present in this windswept lochan. The plants were small and reproducing vegetatively by stolons: small plants were growing at the tips of the stolons and rooting from stolons which were still attached to the parent plant. Some plants were also in bud. Similar dwarf plants with short stolons have been collected at high altitudes on Meall nan Tarmachan for many years, most recently by R. Mackechnie & E. C. Wallace on 8 July 1937 (BM). Meall nan Tarmachan is cited as the highest locality for this species by Dandy & Taylor (unpublished) on the basis of these collections. Wilson’s (1956) upper limit of 3350 feet [1020 m] is based on Bennett (1907), who actually cited an unlocalised record from 3300 feet [1005 m] as the upper limit for this species in Britain. This higher value is best disregarded in the absence of further information. The altitudinal limits in Bennett (1907) are based on specimens he himself had seen, and he perhaps derived this figure from an imprecisely labelled sheet from “near the summit’’ of Meall nan Tarmachan (1043 m). Potamogeton filiformis. Although this species has a northern distribution in Britain, it is primarily a lowland plant which ascends to 350 m in Drumore Loch, Angus, v.c. 90 (Stewart, Pearman & Preston 1994). The only exception is a specimen collected from “alt. near 2500 ft. [760 m]”’ at “lochan above Coire Dhubh Ghalair, Breadalbane”’ by D. A. Haggart on 1 August 1889 (PTH). The identification of this specimen was initially confirmed by Dandy & Taylor in 1939 and subsequently by C.D.P. in 1997. The site, Lochan Coire Dhubhclair, lies at an altitude of 735 m. We revisited it on 27 July 1995: it is a stony lochan with no emergents, although green algae cover the rocks. Small plants of Juncus bulbosus grew around the edge of the loch but we could find no Potamogeton filiformis. There is no reason to suppose that the specimen at PTH is mislabelled (M. Simmons, in litt., 1997), but it is possible that the species was only a transient colonist at this apparently unsuitable site which is 50 km from any other known population. Potamogeton perfoliatus. With P. praelongus in water 20-30 cm deep at the shallow edge of Lochan an Tairbh-uisge, Meall nan Tarmachan, altitude 780 m, NN/591.396, 25 July 1995, CGE (Preston 95/ 5]). Although P. praelongus has been collected at this site before (see below), P. perfoliatus appears to have been overlooked (or perhaps regarded by earlier botanists as too common to collect). This site exceeds the upper altitudinal limits given by Wilson (1956) and Dandy & Taylor (unpublished), both of which are based on E. S. Marshall’s collection from 750 m at Loch Vrotachan, v.c. 92 (Marshall 1893; voucher in CGE), a site where the species was still present in 1996. Potamogeton polygonifolius. Outflow of Lochan an Tairbh-uisge, altitude 780 m, NN/591.396, 25 July 1995. This exceeds the published upper limit of this species, 700 m in Breadalbane (White 1898; Wilson 1956). According to Dandy & Taylor (unpublished), the species occurs “up to about 3000 NOTES 189 ft. [915 m] on Meall nan Tarmachan’’; there are no specimens with precise altitudinal data in the Dandy index, although it details one collection by A. B. Hall from “pools near the summit of Meall nan Tarmachan”’ (July 1892). We failed to find this species at higher altitudes on Meall nan Tarmachan, although it might occur there. Potamogeton praelongus. Flowering plants in water 50-60 cm deep, and a few scattered plants with P. perfoliatus in water as shallow as 20-30 cm, Lochan an Tairbh-uisge, Meall nan Tarmachan, altitude 780 m, NN/591.396, 25 July 1995, CGE (Preston 95/49, 50). This has hitherto been regarded as the highest locality for this species in Britain: it has been collected here since 1881, most recently by A. O. Chater in 1953 (LANC). However, a higher locality was discovered in 1995 by the Scottish Natural Heritage Loch Survey team, who collected P. praelongus at 800 m at Loch Coire Cheap, NN/480.754, in v.c. 97 (voucher specimen confirmed by C.D.P.). References to the occurrence of P. praelongus at higher altitudes on Meall nan Tarmachan are based on a specimen collected by H. N. Dixon & A. H. Vallance in 1893 (Dandy index) on which the altitude of this site is given, rather imprecisely, as 2500-3000 feet [760-915 mJ]. Both Bennett (1903) and Wilson (1956) chose the upper of these two values as the upper limit for the species. Potamogeton x zizii (P. gramineus x P. lucens). Clear water 1 m deep, with Littorella uniflora, Lobelia dortmanna, the aquatic variant of Juncus bulbosus and Myriophyllum alterniflorum, Loch na Craige, altitude 395 m, NN/88.45, 29 July 1995, CGE (Preston 95/64). Wilson (1956) described this as a lowland plant, although there are earlier collections from this site, made by G. Taylor in 1932 and 1933 and by J. W. Clark in 1973. These collections were initially determined as P. lucens by W. H. Pearsall and J. E. Lousley respectively (Dandy index). Sparganium angustifolium. A few plants in a lochan fed by two melting snowpatches, 200 m S.E. of the summit of Beinn Heasgarnich, altitude 1005 m, NN/415.382, and larger populations in a stony lochan 500 m N.E. of the summit, altitude 970 m, NN/418.385, CGE (Preston 95/56), 26 July 1995. Wilson (1956) includes both this species and S. natans in his entry for S. angustifolium, although suggesting (doubtless correctly) that the higher altitude records probably refer to S. angustifolium. None of his entries, however, is for an altitude as high as 1005 m. Druce (1932) cites a record made by P. Ewing at 990 m from Ben Lawers; this puzzling record may refer to the Lawers area as there does not appear to be any suitable habitat at this altitude on Ben Lawers itself. Sparganium natans. Growing with Potamogeton natans in water 30-40 cm deep over peat (with submerged inflorescences), and as tiny plants in liquid mud (with emergent inflorescences), Lochan Achlarich, E. of Beinn Heasgarnich, altitude 650 m, NN/434.380, 26 July 1995, CGE (Preston 95/ 57). As explained under S. angustifolium, Wilson (1956) does not specify an altitudinal limit for this species. This record is higher than any others we have traced. Triglochin palustris. By lochan 500 m N.E. of the summit of Beinn Heasgarnich, altitude 970 m, NN/ 418.385, 26 July 1995. Macvicar (1894) and White (1898) give the upper limit for this species as 855 m in Breadalbane, the value cited by Wilson (1956), although Druce (1932) cites an upper limit of 975 m on Lochnagar. The presence of this species at high altitudes in Britain is not unexpected, as it reaches high latitudes in the Arctic (Hultén & Fries 1986). Utricularia intermedia sensu lato. Shallow water over peat, Lochan Achlarich, E. of Beinn Heasgarnich, altitude 650 m, NN/434.380, 26 July 1995, CGE (Preston 95/59). Also present at the same altitude in a small lochan N. of this site, NN/432.381. This exceeds the upper limit given by White (1898) and Wilson (1956), 550 m in the Breadalbane and 565 m in the Rannoch areas of Perthshire. However, Druce (1932) cites a record made by P. Ewing at 990 m on Ben Lawers, a record so much higher than the others that confirmation is desirable. Nitella flexilis. In lochan fed by two melting snowpatches, 200 m S.E. of the summit of Beinn Heasgarnich, altitude 1005 m, NN/415.382, 26 July 1995, CGE. Forming low sward in water 20- 30 cm deep over soft silt, in absence of vascular plants, smaller lochan S.W. of the summit of Meall nan Tarmachan, altitude 950 m, NN/581.386, 25 July 1995, CGE. Deep water in lochan on peat (but just below rock outcrops), N. end of Coire Heasgarnich, altitude 930 m, NN/420.389, 26 July 1995, CGE. Lochan Coire Dhubhclair, altitude 735 m, NN/497.329, 27 July 1995, CGE. N. F. Stewart now splits this species into the segregates N. flexilis and N. opaca. The specimens from 1005 m on Beinn Heasgarnich can only be identified as N. flexilis sensu lato but N.F.S. has identified the other three plants as N. opaca. Wilson’s (1956) upper limit for the genus, N. flexilis at 545 m, is clearly much too low. 190 NOTES ACKNOWLEDGMENTS We thank Mrs A. V. Pearman for joining us on our visits to all the Breadalbane sites, D. Marden (National Trust for Scotland) and Dr R. A. H. Smith (Scottish Natural Heritage) for permission to collect voucher specimens on Meall nan Tarmachan, M. Simmons for the loan of Haggart’s P. filiformis specimen from PTH, Dr O. Lassiére and Dr D. A. Ratcliffe for providing details of Potamogeton praelongus and Carex rostrata and N. F. Stewart for determining the charophyte specimens. REFERENCES BABINGTON, C. C. (1881). Manual of British botany, 8th ed. John van Voorst, London. BENNETT, A. (1903). Potamogeton praelongus Wulf. in Britain. Journal of botany 41: 165-166. BENNETT, A. (1907). Notes on Potamogeton. Journal of botany 45: 373-377. DANDY, J. E. & TAYLOR, G. (unpublished). The British species of Potamogeton L. Manuscript in Natural History Museum library (catalogued as DF 440/65, 66), London. Druce, G. C. (1932). The comital Flora of the British Isles. T. Buncle & Co., Arbroath. HULTEN, E. & Fries, M. (1986). Atlas of north European vascular plants north of the Tropic of Cancer. 3 vols. Koeltz Scientific Books, K6nigstein. JERMY, A. C., CHATER, A. O. & DAVID, R.W. (1982). Sedges of the British Isles. B.S.B.1. Handbook no. 1, 2nd ed. Botanical Society of the British Isles, London. Macvicar, S. M. (1894). Altitudes reached by certain plants in Mid-Perth. Annals of Scottish natural history 1894: 164-166. MARSHALL, E. S. (1893). Some plants observed in E. Scotland, July and August, 1892. Journal of botany 31: 228— 236. McVeEan, D. N. & RATCLIFFE, D. A. (1962). Plant communities of the Scottish Highlands. Monographs of the Nature Conservancy no. |. Her Majesty’s Stationery Office, London. Moore, J. A. (1986). Charophytes of Great Britain and Ireland. B.S.B.1. Handbook no. 5. Botanical Society of the British Isles, London. PRESTON, C. D. (1988). The Potamogeton L. taxa described by Alfred Fryer. Watsonia 17: 23-35. PRESTON, C. D. & Crort, J. M. (1997). Aquatic plants in Britain and Ireland. Harley Books, Colchester. STACE, C. A. (1991). New Flora of the British Isles. Cambridge University Press, Cambridge. STEWART, A., PEARMAN, D. A. & PRESTON, C. D. comps & eds (1994). Scarce plants in Britain. Joint Nature Conservation Committee, Peterborough. Waite, F. B. W. (1898). The Flora of Perthshire. Perthshire Society of Natural Science, Edinburgh. WILSON, A. (1956). The altitudinal range of British plants, 2nd ed. T. Buncle & Co., Arbroath. C. D. PRESTON I.T.E., Monks Wood, Abbots Ripton, Huntingdon, Cambs., PE17 2LS D. A. PEARMAN The Old Rectory, Frome St Quintin, Dorchester, DT2 OHF COLONISATION BY COCHLEARIA OFFICINALIS L. (BRASSICACEAE) AND OTHER HALOPHYTES ON THE ABERDEEN-MONTROSE MAIN ROAD IN NORTH-EAST SCOTLAND Clumps of Cochlearia officinalis L. (Common Scurvygrass) have been visible on the verges of the A92 road south of Stonehaven since the 1980s, and possibly earlier. Because halophytes have been increasing elsewhere in Britain along trunk roads in the last 20 years (Scott & Davison 1982; Scott 1985; Leach 1994), a similar spread could be occurring in N.E. Scotland. We therefore made a car- borne survey of the coastal main road in Kincardineshire (v.c. 91) in late:-April and early May 1997, driving slowly along its whole length in the vice-county looking for flowers of halophyte species. On finding plants we assessed their distribution pattern across the road verges. For the survey we searched between the North Esk bridge just north of Montrose and the Dee bridges in Aberdeen, with extra traverses on key sections. We also checked the branches of the coast road leading into Stonehaven (designated A92 until the mid 1980s when Stonehaven was by-passed) and the A956, which gives a second entry to Aberdeen further east than the main road (Fig. 1). From Aberdeen to the south end of the Stonehaven by-pass, the road is now a dual carriageway and numbered A90; NOTES 19] + Population <10 clumps @ Population 10—100 clumps Aberdeen extending 5-200 m B® Population >100 clumps extending c. 1 km FiGuRE |. Distribution of Cochlearia officinalis on the Aberdeen — Montrose main road in spring 1997. from the Stonehaven by-pass to Montrose the road is a single carriageway apart from a short section near Gourdon (location 5, Fig. 1). Clumps of Cochlearia officinalis were recorded in eight locations, and Armeria maritima (Miller) Willd. (Thrift) in three locations (Table 1). Although we would have missed any non-flowering plants of A. maritima and C. officinalis, and perhaps also some small flowering clumps if obscured by tall grasses, we are confident that major colonisation has occurred on only two sections of road near Dunnottar (locations 3 and 4, Table 1) and Johnshaven (locations 7 and 8). At three other locations only single clumps of C. officinalis were seen. The A. maritima colonies were all small, although the size of the clumps indicates that they have been established for several years. Plantago maritima L. (Sea Plantain) and Spergularia marina (L.) Griseb. (Lesser Sea-spurrey) were also found, each at one location (locations 5 and 3 respectively), when we were examining the verges on foot. They probably grow in more places along the A90/A92, but would not have been noticed as they flower later than A. maritima and C. officinalis. No plants of Cochlearia danica L. (Danish Scurvystass) were seen in this survey, but in a check on the inland stretch of the A90 in Kincardineshire (v.c. 91) on 20 May 1997, four small clumps were found near Laurencekirk (NO/701.699-726.720). This road is a dual carriageway running W.S.W. from the Stonehaven by-pass (NO/861.843) to the vice-county boundary at North Water Bridge (NO/651.661); it was searched less thoroughly than the coastal main road and at greater speed, and no other halophytes were observed. The distribution of C. officinalis within the colonies on the coast road was somewhat patchy, and at Lauriston the clumps appeared to be confined to one side of the road (Table 1). Across the verges there was a Clear pattern of greater frequency of clumps at the road edge. However, this was much less marked than for C. danica on the M6 and M74 in northern England and southern Scotland (personal 192 NOTES TABLE 1. DETAILS OF OCCURENCES OF ARMERIA MARITIMA (Am) AND COCHLEARIA OFFICINALIS (Co) ON THE ABERDEEN-MONTROSE MAIN ROAD IN SPRING 1997. ALL LOCALITIES ARE IN KINCARDINESHIRE (V.C. 91) Grid Population reference size Spread Locality (all NO) Species (no. clumps) (m) Position 1. Loirston 931.001 Co ] <| East verge 2. Bourtreebush 910.955 Co 1 <1 Central reservation 3. Dunnottar 872.837- Co c. 2000 2700 Both verges 873.810 4. Catterline 866.798-— Co c. 20 100 Both verges 867.799 5. Gourdon 824.711 Am 1 0-4 (0-15—)0-2—0-35(—0-4)) <0-25 length of blade Margin of upper sheath-opening in red-dotted red-dotted yellowish or herbarium specimens yellowish brown Aerenchyma in stem + . . Depth of substomatal cavities (um) 6-7 8-18 20-26 Stomatal length (um) 42-51 39-46 36:5—40-5 Perianth bristles papillose smooth Excretory cells in old herbarium + + - specimens Fruiting + - + + : present, - : absent altitude and latitude) failed to realise that their common plant is not 7. cespitosum subsp. cespitosum, but is subsp. germanicum. Clearly, however, Fernald (1921) regarded Scirpus cespitosus L. as being identical with Palla’s Trichophorum germanicum, but considered the latter name unnecessary in view of Roth’s S. cespitosus B nemorosus (Roth 1789). However, I do not think that Roth’s description identifies the Atlantic-Subatlantic plant. The crux of the matter concerns whether Scirpus cespitosus L. refers to the circumpolar or the Atlantic-Subatlantic plant. The European nomenclature is based on the former, while Fernald suggested that Linnaeus had intended it to be the Atlantic-Subatlantic plant. There is a specimen (Sheet 71.8) in LINN, but it is believed that this could not have been in Linnaeus’ possession until after 1753 and so cannot be considered as an original element for this name, even though it bears the Linnean annotation “‘cespitosus” (N. J. Turland, pers. comm., 1994). However, the specimen, Herb. Linn. No. 20 (LAPP), is almost certainly the circumpolar plant and can therefore be designated as the lectotype. This will fix the application of the name Scirpus cespitosus L. to the circumpolar taxon, 1.e. T. cespitosum L. subsp. cespitosum as it is currently understood in Britain and the rest of Europe. In the present paper I shall continue to follow tradition by using the epithet germanicum for the Atlantic-Subatlantic plant. Fernald’s nomenclature has been followed by authors of North American Floras, e.g. Beetle (1947) and Scoggan (1978), as well as in Japan (Koyama 1958). However, Hultén (1962) and Hegi (1966) have reaffirmed acceptance of the European nomenclature. DISTRIBUTION In Clapham et al (1952, 1962) subsp. cespitosum is stated to be known only from Ingleborough and Ben Lawers, but Clapham et al. (1987) say “rare and its distribution is imperfectly known”. However, in Clapham et al. (1981) subsp. germanicum is given, quite erroneously, as “rare” and subsp. cespitosum as “common’’. In Stace (1991) and Kent (1992) subsp. cespitosum is omitted, evidently because no clear evidence of its occurrence in the British Isles could be found. Sell & Murrell (1996) state “All our plants seem to be subsp. germanicum, for although there are records of subsp. cespitosum they have not been substantiated, but plants intermediate between the subspecies have been recorded in widely scattered localities.”” Nevertheless, in Stace (1997) subsp. cespitosum has been reinstated. In Northumberland subsp. germanicum is the common plant, growing in acidic peat in moorland and it appears to tolerate a variety of habitats of varying degree of wetness. However, as I first recognised in May 1988, there is also another plant in Northumberland, which is characteristic of the Border Mires (raised mires), but which also occurs on other Sphagnum mires. The distribution TRICHOPHORUM CESPITOSUM NOTHOSUBSP. FOERSTERI 211 of this is given in Swan (1993), where it was named T. cespitosum subsp. cespitosum, although queried as being possibly a hybrid. The sheath character of this plant is closer to that of subsp. cespitosum than to subsp. germanicum. This plant from Northumberland, which has never been found to produce mature fruits, is now believed to be a sterile hybrid between subsp. cespitosum and subsp. germanicum, as demonstrated by morphological and anatomical work described in this paper, and confirmed by work using isozyme electrophoresis carried out by P. M. Hollingsworth (Hollingsworth & Swan 1999). A proliferous form also occurs. Since 1988 I have extensively investigated 7. cespitosum in Northumberland, but have found subsp. cespitosum at only four sites (all in v.c. 67), all less strongly acidic, or base-rich, or less stagnant than those where subsp. germanicum or the hybrid occur. AIMS OF THE WORK 1. To assess critically the values of the various characters which have been used in attempting to identify the three taxa, and where appropriate relating such characters to geographical range. 2. To establish the plant of the Border Mires as being a hybrid. 3. To determine the distributions and habitats of subsp. cespitosum and the hybrid in Britain and Ireland and (so far as possible) elsewhere in N. W. Europe, and in particular to resolve the apparent anomaly that in Norway subsp. cespitosum is the usual plant of raised mires, whereas in Britain and Germany it is the hybrid. 4. To investigate the distributions and habitats of the proliferous plants. IDENTIFICATION GENERAL Nomenclature of vascular plants follows Kent (1992). Subsp. germanicum is the common plant, forming large, dense tufts, growing in acidic peat, often with Calluna vulgaris, Erica tetralix, Juncus squarrosus and Molinia caerulea. The broad lower glume sometimes overtops the comparatively large spikelet, but these characters are not reliable for identification. The hybrid is less often so densely cespitose as the above and grows in active Sphagnum bogs (blanket, raised or valley mires), often with Andromeda polifolia, Calluna vulgaris, Drosera rotundifolia, Erica tetralix, Eriophorum angustifolium, E. vaginatum, Narthecium ossifragum, Vaccinium oxycoccos and, occasionally, also Drosera longifolia and Rhynchospora alba. Subsp. cespitosum has very small spikelets. In Northumberland it has not been seen in large, dense tufts; it sometimes occurs in base-rich habitats and never under such strongly acidic or stagnant conditions as the hybrid. Characters which have been used to separate the three (non-proliferous) taxa are discussed below (followed by a description of proliferous forms) and are summarised in Table 1. Measurements relate to material from Northumberland. UPPER SHEATH In subsp. germanicum the uppermost sheath has an oblique opening 2—3(—5) mm long and 1 mm broad, the blade being up to 2 x as long as the opening (see key in Foerster 1963). In subsp. cespitosum the opening of the sheath is suborbicular, c. 1 mm in diameter, with the blade 5—10 x as long as the opening. Note that Fremstad & Skogen (1978) give sheath-opening up to 1-5 mm, which overlaps with measurements on the hybrid. It should be mentioned that in Hegi (1909) the drawings of the sheaths of the two subspecies (Fig. 186, p. 24) have been reversed, although the descriptions (on p. 25) are correct. Even in the third edition of this excellent work (Hegi (1966), Fig. 24, p. 37) this error still remains, which perhaps shows how neglected is T. cespitosum [Cinderella of the Cyperaceae!]. However, this character is correctly depicted in Clapham et al. (1987) and in Hess et al. (1967). Unfortunately in Stace (1997) the sheath characters of the two subspecies have also been reversed. In the hybrid, the sheath-opening is 1-1-5(—2) mm long, with the blade 3-7 x as long as the opening. 22 G. A. SWAN Length of sheath-opening (mm) 0.1 0.2 0.3 0.4 0.5 0.6 0.7 Ratio of length of sheath-opening to length of blade Ficure |. Scatter diagram of length of sheath-opening (mm) against ratio of length of sheath-opening to length of blade for Trichophorum cespitosum (Northumbrian, non-proliferous, herbarium material). O subsp. cespitosum, ®@ subsp. germanicum, e hybrid. Fremstad & Skogen (1978) showed diagrammatically the relationship between the length of the sheath-opening and the ratio of sheath-opening to blade length for subsp. cespitosum and germanicum. Figs | & 2 show this relationship in material from Northumberland, including the hybrid, for non-proliferous and proliferous material, respectively. Each point on the diagrams represents the average of three measurements on the same specimen. To see and measure the sheath-opening, it is convenient to first pull out the stem from the sheath. Palla (1897) and various other authors have contrasted the sheath-opening of subsp. germanicum with that of subsp. cespitosum as being loose-fitting, with the scarious margin dotted with red or rusty brown, as opposed to close-fitting with the margin yellowish or yellowish brown. However, the first character is relevant only to herbarium material and is likely to be confusing on fresh material (Foerster 1963). Possibly, the presence of aerenchyma in subsp. germanicum results in shrinkage of the stem on drying, and hence the appearance of a loose-fitting sheath. The rusty colour is usually very obvious in herbarium specimens of subsp. germanicum and often also in those of the hybrid, so its absence can help to confirm subsp. cespitosum. STEM ANATOMY Cross-sections of fresh stems or dried stem material (the latter soaked in hot water), were cut with a razor blade and examined at x 30. In subsp. germanicum aerenchyma is clearly seen (Fig. 3d), whereas in subsp. cespitosum it is completely absent (Fig. 3a, 3b). In many specimens of the hybrid, aerenchyma is also absent, although in some there are insular spaces, but not continuous channels (Fig. 3c). The clearest distinction between the three taxa is seen in the substomatal cavities, best seen at higher magnification (x 400). The stomata run in rows, lengthwise along the stem, and beneath each row of stomata lie the substomatal cavities, which in the case of subsp. cespitosum are lined by thick-walled, cutinised cells without chlorophyll. These rows are distributed around the circumference of the stem. In cross-sections of the stem, the substomatal cavities are seen around TRICHOPHORUM CESPITOSUM NOTHOSUBSP. FOERSTERI 213 4.0 oe ro) Length of sheath-opening (mm) = N (o>) (2) 0.1 0.2 0.3 0.4 0.5 0.6 0.7 Ratio of length of sheath-opening to length of blade Ficure 2. Scatter diagram of length of sheath-opening (mm) against ratio of length of sheath-opening to length of blade for Trichophorum cespitosum (Northumbrian, proliferous, herbarium material). O subsp. cespitosum, @ subsp. germanicum, e hybrid the periphery, and each with a guard-cell at the outside. The number of cavities through which a particular section cuts varies, but under favourable circumstances it may be up to 12 (some of them double) in subsp. cespitosum. The stem cross-sections of the two subspecies are illustrated in Broddeson (1912), and the arrangement of the substomatal cavities in a length of stem as well as the construction of the cavities and guard-cells (in T. alpinum) is illustrated in Westermaier (1881). Illustrations of the cavities in subsp. cespitosum are given in Montfort (1918) and Metsdvainio (1931). There is a good drawing (with scale) of a cross-section showing a substomatal cavity in Firbas (1931), p. 485. Montfort (1918) showed that, unlike other species of raised mires, Eriophorum vaginatum and T. cespitosum both have substomatal cavities as described above, the physiological action of which may be to lower rates of transpiration. These cavities increase stomatal resistance and form physiological barriers to transpiration (diffusion pathway lengthened). As a consequence of this structure the plants are anatomically xeromorphic. Both species flower very early, often under semi-frozen conditions, so that the plants may suffer “physiological drought”. Eriophorum latifolium flowers three to four weeks later than E. vaginatum, after the peat has thawed completely and lacks this xeromorphic character. In T. cespitosum subsp. cespitosum the xeromorphic anatomy is extremely strongly developed, not only in plants from raised mires, but also in other mire systems. The suggestion by Ellenberg (1988) and Lusby & Wright (1996), that the supposed xeromorphic character of bog plants is not true xeromorphism, but is peinomorphism, should perhaps refer only to evergreen ericaceous species. Montfort (1918) suggested that the xeromorphic character of Eriophorum vaginatum and T. cespitosum subsp. cespitosum dates back to the glacial period. In subsp. germanicum (Atlantic-Subatlantic) the substomatal cavities are very small, their radial diameter (6—7 tm) being only half that of the guard-cells, while in subsp. cespitosum (circumpolar) they are large, with radial diameter (20-26 um) up to twice that of the guard-cells. In a good section 214 G. A. SWAN ee FiGure 3. Cross-sections of stem of Trichophorum cespitosum subsp. cespitosum from below Schwarzsee, near Zermatt, Switzerland (a and e), of T. cespitosum subsp. cespitosum from Blackheugh End, S. Northumberland (v.c. 67) (b and f), of 7. cespitosum nothosubsp. foersteri from Great Wanney Crag Moss, S. Northumberland (v.c. 67) (c and g) and of T. cespitosum subsp. germanicum from S. Northumberland (v.c. 67) (d). The scale bars represent 0-2 mm (a—d) and 20 um (eg). TRICHOPHORUM CESPITOSUM NOTHOSUBSP. FOERSTERI 215 50 3 = reas fo) e a is = oe 40 ~ Selim 2015) POS, 40:5) 0:6), 027 25038) 0:94 Overt of.2 5 113..1.4 Ratio of length of sheath-opening to length of blade Ficure 4. Scatter diagram of stomatal length (um) against ratio of length of sheath-opening to length of blade for Trichophorum cespitosum (Northumbrian, non-proliferous, herbarium material) O subsp. cespitosum, @ subsp. germanicum, e hybrid. of subsp. cespitosum, under the microscope these substomatal cavities stand out as a ring of circular (or sometimes slightly squarish) “windows” (some double) around the section. In specimens collected below Schwarzsee, near Zermatt, in Switzerland, at c. 2500 m, the cavities were very large and numerous (Fig. 3a, 3e); in material from Northumberland they were somewhat smaller and less numerous (Fig. 3b, 3f), although the plants were taller than those from the Alps. In the hybrid the substomatal cavities may be seen as circular (of smaller diameter than in subsp. cespitosum) (Fig. 3c, 3g), but more often as oval openings, their depth (towards the centre of the section) (usually 14—17 um) being less than their width (along the wall of the section) (up to c. 28 um). The number of cavities seen in a section of the hybrid is usually less than in subsp. cespitosum. ) The aerenchyma of subsp. germanicum is presumably relevant to growth under wet Atlantic- Subatlantic conditions. Examination of a stem-section is an excellent means of separating the three taxa. STOMATAL LENGTH Fremstad & Skogen (1978) reported the stomatal lengths as being 48-62 + 2-39 and 42-29 + 2-63 tum in subspp. germanicum and cespitosum, respectively. They showed diagrammatically the relationship between stomatal length and the ratio of the lengths of sheath-opening and blade. In the present work herbarium specimens of T. cespitosum from Northumberland were soaked in hot water before the epidermis was stripped. The lengths of 20 stomata from each specimen (from approximately the middle of a stem) were then measured (at x 400) and the average stomatal length was plotted against the ratio of lengths of sheath-opening to blade, as shown in Figs 4 & 5 for non-proliferous and proliferous material respectively. In agreement with the Norwegian authors, the shortest stomata were found in subsp. cespitosum and the longest in subsp. germanicum; the hybrid occupied an intermediate position. In a small number of specimens it was found that the average stomatal length varied considerably in different parts of the stem. 216 G. A. SWAN Stomatal length (um) 01° 012, 03 04 05 06 O07 08.09 1.0 4.4 23s Ratio of length of sheath-opening to length of blade Ficure 5. Scatter diagram of stomatal length (um) against ratio of length of sheath-opening to length of blade for Trichophorum cespitosum (Northumbrian, proliferous, herbarium material) @ subsp. germanicum, e hybrid. FURROWED STEM Fremstad & Skogen (1978) stated that whereas the stems of subsp. germanicum were furrowed, those of subsp. cespitosum were smooth. I noticed that the stems of subsp. cespitosum were indeed often only shallowly grooved, while those of the hybrid were usually deeply grooved. However, this character is subject to seasonal variation and is of little value. PERIANTH BRISTLES In subsp. germanicum the perianth bristles are papillose. In subsp. cespitosum they are smooth, except sometimes near the apex. In agreement with Palla (1897) who recorded this character, I did not find this to be a ngorous means of separation. There are excellent SEM photographs of bristles of the two subspecies in Fremstad & Skogen (1978). BASAL SHEATHS Clapham et al. (1987) and De Filipps (1980) describe subsp. germanicum as having basal sheaths scarcely shining, whilst subsp. cespitosum has basal sheaths shining. Sell & Murrell (1996) likewise give “dull” and “shining”, respectively. It is true that when one pulls the plant (e.g. the hybrid) out from a cushion of Sphagnum, the basal sheaths are usually shining, but I cannot regard this as a satisfactory character for identification. According to Ostenfeld & Gréntved (1934), the old leaf sheaths in subsp. germanicum are “pale-brown, often dark from decaying matter, hardly shining”, while in subsp. cespitosum they are “bright-brown and shining”. This agrees in the main with what is observed in Northumberland, but the difference may merely reflect the different sheath environments (i.e. peat as opposed to living Sphagnum) rather than being a character of the subspecies. EXCRETORY CELLS Palla (1897) and Fremstad & Skogen (1978) mentioned the presence and absence of excretory cells (red-brown) in the assimilatory tissue of subsp. germanicum and cespitosum, respectively. In TRICHOPHORUM CESPITOSUM NOTHOSUBSP. FOERSTERI 217 FicureE 6. Spikelet of proliferous Trichophorum cespitosum nothosubsp. foersteri. material from Northumberland these coloured cells were observed only in herbarium specimens which had been kept for at least seven years. Old specimens of either subsp. germanicum or the hybrid usually show these red cells, whereas they are absent in subsp. cespitosum, so this can be a useful confirmation in the case of herbarium specimens. FRUITING Flowering begins in May and, in the case of subsp. germanicum, well developed fruits (to c. 2 mm) are formed; spikelets with fruits (containing hard nutlets) and perhaps a few glumes can still be found in August. Subsp. cespitosum also produces smaller, short, broad heads of dark fruits. However, in the case of the hybrid, many of the tufts fail to produce fruit at all, and by early July the tops of their stems are more or less bare; in other tufts, development from the flowering to the fruiting stage proceeds and by the beginning of July some small, green, apparently sterile “fruits” (maximum length 1 mm) are present, but soon the glumes are shed. Hard nutlets have never been found. The flowers of 7. cespitosum are generally stated to be hermaphrodite, e.g. Clapham et al. (1987), Sell & Murrell (1996). However, the last sentence in the account of the species in Hegi (1909) states that some tufts have only protogynous, hermaphrodite flowers, while others have female and male flowers. The occurrence of some tufts with purely male flowers could account for the stems of some Northumberland material becoming bare at the top by early July. Fruits of subsp. cespitosum from the Arctic have been described by Polunin (1959) and in North American plants by Beetle (1947), yet there seems to be a dearth of information on the occurrence or failure of fruiting of subsp. cespitosum in Germany and Scandinavia. However, Foerster (1963), on an excursion to the Hohes Venn (near the border between Germany and Belgium) described finding subsp. cespitosum on the southern slope of the Pannensterz, where it was growing sparingly among much subsp. germanicum and Molinia caerulea. He stated that on 27 July 1962 the fruits of the subsp. cespitosum there had already fallen and that the subsp. cespitosum there was not more delicate than the subsp. germanicum, which he suggested might be the result of the stems continuing to lengthen after the ripening of its fruit. In Northumberland the stems of the hybrid continue to lengthen during July and August, even though their tops are bare; and Foerster’s observations seem entirely consistent with his plant in the Hohes Venn being the hybrid, rather than subsp. cespitosum and having failed to fruit, rather than having dropped its fruits, as stated. PROLIFEROUS T. CESPITOSUM In raised mires in Northumberland, in addition to the hybrid of T. cespitosum subsp. germanicum with subsp. cespitosum, a proliferous (“viviparous”) form of this hybrid also occurs. From the middle of June onwards this can easily be distinguished at sight from the common form, the spikelets being broader and variegated with off-white and green (Fig. 6). This is reminiscent of 218 G. A. SWAN TABLE 2. SPECIES GROWING WITH TRICHOPHORUM CESPITOSUM SUBSP. CESPITOSUM AT BLACKHEUGH END AND GOWANY KNOWE (V.C. 67) Blackheugh End Gowany Knowe Cardamine pratensis Carex dioica Carex flacca Carex hostiana Carex hostiana Carex lasiocarpa Carex panicea Carex limosa Carex pulicaris Carex panicea Carex rostrata Carex pulicaris Carex viridula subsp. brachyrrhyncha Carex rostrata Drosera rotundifolia Carex viridula subsp. brachyrrhyncha Eleocharis quinqueflora Dactylorhiza incarnata Equisetum palustre Dactylorhiza maculata subsp. ericetorum Erica tetralix Drosera rotundifolia Eriophorum angustifolium Menyanthes trifoliata Eriophorum latifolium Pedicularis palustris Eriophorum vaginatum Phragmites australis Menyanthes trifoliata Potentilla palustris Pinguicula vulgaris Ranunculus flammula Polygala serpyllifolia Potentilla erecta Salix repens Selaginella selaginoides Taraxacum faeroense Triglochin palustre Vaccinium oxycoccos Viola palustris Festuca vivipara. The lower flowers in the spikelets develop to form small, sterile “fruits”, whereas the upper flowers proliferate to give green plantlets. Also, in all the flowers, the bristles are replaced by a membranous (petaloid) perianth. A particular tuft is either proliferous or non- proliferous, and the proliferous plants usually form smaller tufts than the non-proliferous ones and retain their glumes much longer, so that they can be found up to early September. Even so their “fruits” do not develop much beyond that of the non-proliferous hybrids, and it is unlikely that they ever produce viable seed. Spikelets of the proliferous hybrid, cut off and then planted in peat in late summer, failed to root satisfactorily, but this does not imply that the propagules fail to do so naturally, when they would fall into the Sphagnum bog. Although plants of the non-proliferous hybrid grew well in peat in pots, plants of the proliferous hybrid grew less well, rarely producing proliferous spikelets, and sometimes even dying during the winter. In some mires, only a few tufts of the proliferous form have been found, whereas in others perhaps up to 20% of the tufts are proliferous. The proliferous tufts tend to be clustered together in the same area of the mire. There is also a proliferous form of subsp. germanicum, but this seems to be rare in Northumberland. It was found on 15 August 1995 near Hareshaw Head at NY/856.884, 335 m (v.c. 67), the site probably being a remnant of a raised mire which had been partly drained. HABITAT OF NON-PROLIFEROUS T. CESPITOSUM The hybrid grows among living Sphagnum, while subsp. germanicum has less strict habitat requirements and grows in other types of peat bog, some of which were formerly Sphagnum- dominated and identical with many surviving Sphagnum mires. In Northumberland the hybrid sometimes grows in a rather bare area of bog, accompanied only by Sphagnum tenellum. Subsp. cespitosum in Northumberland has been found only at the margins of raised or valley mires, where there is some water-movement and base enrichment, never actually within the mire. TRICHOPHORUM CESPITOSUM NOTHOSUBSP. FOERSTERI 219 LOCALITIES OF SUBSP. CESPITOSUM IN NORTHUMBERLAND Subsp. cespitosum, with mature fruits, has been found by G.A.S. at four sites in South Northumberland (v.c. 67). See Table 2. 1. Near Blackheugh End, NY/826.915 (330 m) on 17 July 1995. Here there is a more-or-less flat area of Sphagnum mire, with Drosera rotundifolia, Erica tetralix, Narthecium ossifragum and hybrid T. cespitosum. Running through this, in an approximately SW/NE direction, is a channel in which grow the species given in Table 2, many of which favour base-rich habitats. It is in this channel that 7. cespitosum subsp. cespitosum grows and where it was first recognised in Northumberland. This site is by the Pennine Way and many walkers pass it, especially during the summer. 2. Gowany Knowe, NY/727.787 (280 m) on 20 July 1996. Gowany Knowe Moss is one of the Border Mires, a raised mire with Carex magellanica, Drosera rotundifolia, Narthecium ossifragum, Vaccinium oxycoccos and hybrid T. cespitosum (including proliferous material). However, T. cespitosum subsp. cespitosum does not grow in the raised mire, but by the side of a tiny streamlet at the margin of the mire, only a very short length of which still remains outside the afforested area. This small habitat also has the species listed in Table 2; it is probably rarely visited, although the Moss is a reserve of the Northumberland Wildlife Trust. 3. Head of Bucklake Sike, NY/69.94 (420 m). A specimen was collected by G.A.S. on 28 June 1990 and this appeared to be subsp. cespitosum, which was confirmed by the collection of a fruiting specimen on 31 August 1996. Here there is a raised mire containing Betula nana, Calluna vulgaris, Carex rostrata, Molinia caerulea, Polygala serpyllifolia, Potentilla erecta, Salix repens and Vaccinium oxycoccos. The lower end of this mire is drained by a tiny streamlet containing Menyanthes trifoliata, Carex panicea and C. pulicaris and this is where T. cespitosum subsp. cespitosum grows. Further downstream Carex limosa appears and also Hammarbya paludosa, although the latter has not been seen since 1973. Still further downstream is Potamogeton polygonifolius. This is a remote site on land owned by the Duke of Northumberland. The proliferous hybrid is also present. 4. Muckle Moss, 4 September 1996. This is a valley mire, which contains the hybrid. Subsp. cespitosum was not found in the mire itself, but in the margin (“lagg’”’) at NY/796.670 (230 m), with Calluna vulgaris and Salix repens. This is a National Nature Reserve and the lagg has been reported to contain Sphagnum balticum, S. majus and S. riparium. Strangely, it is stated in Ratcliffe (1977) that T. cespitosum is conspicuously absent in Muckle Moss. In addition, I investigated many base-rich flushes on peat in Northumberland, each containing several of the species listed in Table 2, but found no Trichophorum in any of these. HABITATS IN SCANDINAVIA According to Fremstad & Skogen (1978), subsp. cespitosum is first and foremost a bog species which occurs throughout Scandinavia (Sjérs 1950). Osvald (1923) has described the occurrence in Sweden of raised mire communities, which appear to be very like those in Northumberland, referred to later in the present paper as Erica-Sphagnum mire (National Vegetation Classification M18, Rodwell 1991). : According to Fremstad & Skogen (1978), when subsp. germanicum appears in a mire it is usually the result of human influence in the form of peat-digging, burning, grazing or trampling. It also occurs where peat growth has stopped and where it is periodically drier than the normal bog. So, in Scandinavia, subsp. germanicum grows commonly with a range of species which are not normally associated with raised mires, such as Carex binervis, Cornus suecica, Deschampsia flexuosa, Festuca vivipara, Luzula multiflora and Polygala serpyllifolia in addition to the normal bog species. Subsp. germanicum also occurs on thin peat and at the edge of bogs. The moss cover is poorly developed, the commonest species being Sphagnum compactum, S. molle, S. strictum and S. tenellum and other species which thrive where peat growth stagnates. Subsp. germanicum also grows on thin peat on paths and trampled places, where peat erosion occurs. According to Fremstad & Skogen (1978), subsp. germanicum has never been found in west- or mid-Norway on intact peat bog. i 220 G. A. SWAN COMPARISON OF HABITATS IN NORTHUMBERLAND AND SCANDINAVIA From the above, it is seen that subsp. cespitosum is the plant which grows in intact peat bogs in Scandinavia, while in Northumberland the hybrid is the characteristic plant in raised mires. There may be two possible explanations for this apparent anomaly. 1. Perhaps subsp. cespitosum was once the plant of mires in both Scandinavia and Britain; but with the post-glacial amelioration of the climate, in the British mires it became too warm for subsp. cespitosum, so that it could not compete successfully with the hybrid, which gradually replaced it. In Scandinavia, the high altitude and/or latitude of the mires allowed subsp. cespitosum to thrive. According to Clapham et al. (1987), T. cespitosum is absent from base-rich soils in the British Isles, and this does indeed seem to be generally true for subsp. germanicum. However, Hegi (1966) and Hess et al. (1967) mention calcareous or base-rich habitats for subsp. cespitosum in particular. T. cespitosum is presumably wind-pollinated and in Northumberland the flowering seasons of the two subspecies probably overlap. Perhaps subsp. cespitosum has only escaped extinction through hybridisation by surviving in marginal areas of raised mires, where conditions are sufficiently base-rich to inhibit the growth of subsp. germanicum. 2. Possibly the hybrid does in fact occur in at least some mires in Scandinavia, but this has not yet been recognised. In the diagram of sheath-opening against the ratio of sheath-opening to blade length, given (as Fig. 1 on p. 136) in Fremstad & Skogen (1978), the group of records around 1-3/0-3 could represent the hybrid rather than subsp. cespitosum (cf. this paper, Fig. 1) and the same could be said of the diagram of stomatal length/ratio of sheath-opening to blade length, given (as Fig. 4 on p. 139), where the group of records around 44/0-2 could represent the hybrid (cf. this paper, Fig. 4). Also in their diagram (Fig. 3 on p. 138), the distribution of stomatal lengths for subsp. cespitosum looks as though it might more probably represent a superposition of two taxa (i.e. subsp. cespitosum and the hybrid). Moreover, specimens in BM from Norway, Jamtland, marshy ground, 750 m, 2 August 1958, H. Smith and in K, Norway, Tromsé Island, in a peat bog, July 1921, V. Summerhayes, are evidently the hybrid. The northernmost locality yet recorded for subsp. germanicum is in Norway, Lofoten (Sortland 1992), somewhat south of Troms6. There are specimens of what appear to be subsp. germanicum from Greenland (K); and from Iceland, 17 July 1876, C. Ostenfeld (WU). Another specimen from Iceland, collected by N. Polunin (BM) seems to be the hybrid. In the intact raised mires of the Northumbrian Border Mires, the hybrid is the characteristic and perhaps the only Trichophorum present; the community in which it grows is evidently Erica- Sphagnum mire [Erica tetralix-Sphagnum papillosum raised and blanket mire, Sphagnum magellanicum-Andromeda polifolia sub-community (M18)]. According to Fremstad & Skogen (1978), when subsp. cespitosum grows in coastal areas of Norway it flowers much earlier than subsp. germanicum, thus preventing hybridisation. However in areas of higher altitude and latitude the growing season is short, so that the flowering seasons of the two subspecies overlap, making hybridisation possible. In these areas morphologically intermediate forms have in fact been observed. Although most of the plants from Northumberland fall fairly clearly into one of the three groups (i.e. subsp. cespitosum, subsp. germanicum and hybrid) a few have been found with mixed characters. These few have been omitted from Figs 1 & 4; they may correspond to Fremstad & Skogen’s (1978) intermediate forms. The hybrid sometimes has good pollen and it is possible that this could cause introgression into subsp. germanicum. DISTRIBUTION DISTRIBUTION IN NORTHUMBERLAND In Northumberland subsp. germanicum mostly occurs from sea-level to 700 m. The large area of The Cheviot (815 m) above 700 m is virtually free from Trichophorum, although on 21 June 1995 I found two small tufts of subsp. germanicum on the summit plateau at 800 m. This absence may be a result of the long snow-cover there. In Northumberland the hybrid has been found at altitudes between 215 and 660 m and subsp. cespitosum from 230 to 420 m; the upward limits may perhaps TRICHOPHORUM CESPITOSUM NOTHOSUBSP. FOERSTERI 221 TABLE 3. MIRES IN SOUTH NORTHUMBERLAND (V.C. 67) WITH TRICHOPHORUM CESPITOSUM NOTHOSUBSP. FOERSTERI No. Name of mire V..C. Grid reference Altitude (m) l Gowany Knowe Moss 67 NY/730.788 280 2 Felecia Moss 67 NY/721.775 310 3 Harelaw Moss 67 NY/757.771 P25) 4 Haining Head Moss 67 NY/714.748 260 5 Wedges Rigg Moss 67 NY/712.742 260 6 Hummell Knowe Moss 67 NY/705.714 250 7 Bell Crag Flow 67 NY/763.721 310 8 Coom Rigg Moss 67 NY/689.795 320 9 Muckle Samuel’s Crags Moss 67 NY/678.789 300 10 Grain Heads Moss 67 NY/744.735 280 11 Butterburn Flow 70 NY/662.761 280 12 The Flothers 67 NY/699.763 290 13 Limy Sike Moss 67 NY/696.770 280 14 Ottercops Moss 67 NY/948.895 310 15 Drowning Flow 67 NY/760.975 405 16 Horse Hill Moss 67 NY/765.790 250] 17 Hobb’s Flow 67 NY/569.902 380 18 Great Wanney Crag Moss 67 NY/938.834 ils) 19 Pundershaw Moss 67 NY/775.792 245 20 Sweethope Moss 67 NY/944.817 250 21 Peterstone Flow 67 NY/980.918 320 22. Towey Moss 67 NY/734.555 470 23 Falstone Moss 67 NY/708.860 250 24 Muckle Moss 67 NY/796.670 230 25 The Lakes 67 NY/740.773 285 26 Blackaburn Lough Moss 67 NY/765.795 265 27 Crane Moss 67 NT/911.034 250 28 Beldon Cleugh Moss 67 NY/917.504 365 “Site 16 was searched for Trichophorum sp. but none was found. be due to the absence of Sphagnum mires at higher altitude. The hybrid occurs in most intact raised mires in Northumberland, notably the Border Mires (see Lunn & Lunn 1976), but can also be found in other places of higher altitude where Sphagnum cover is less continuous, e.g. near the summit of Windy Gyle (NT/855.152) at 600 m, where there is also some Narthecium ossifragum, or Scotsman’s Knowe (NT/904.191) at 660 m. I also found the hybrid in 1995, just over the border (in Cumberland, v.c. 70) on the N.W. slope of Cold Fell (NY/60.56) at 520 m, growing not in Sphagnum, but in Dicranum scoparium. The proliferous form of the hybrid was first noticed on Haining Head Moss (Table 3, Mire 4) on 13 June 1992, although an indeterminate specimen of a proliferous plant, collected on Harbottle Moor (v.c. 67) on 12 August 1934 by G. W. Temperley was later found in HAMU. During the summers of 1993 and 1994 I sought the proliferous hybrid in the 28 mires, thought to be relatively undisturbed, listed in Tables 3 & 4, as well as elsewhere in Northumberland. All these mires were found to contain Drosera rotundifolia, Narthecium ossifragum and Vaccinium oxycoccos. Mires 1—23 are essentially of the blanket or raised mire type, while 24—28 are more of the valley mire type. Mires 1-16 also contain Andromeda polifolia, hybrid Trichophorum and proliferous hybrid Trichophorum, with the exception of Mire 16, which lacks any Trichophorum at all and although it was in the same area and was otherwise similar to many others, it had been ploughed up by the Forestry Commission around 1949, but never planted with trees. Mire 23 is very dry and lacks Andromeda polifolia and the proliferous Trichophorum. The proliferous Trichophorum hybrid was not found in any of the valley mires, i.e. Mires 24—28, all of which contain the non-proliferous hybrid. Rose (1953) in comparing lowland British valley G. A. SWAN 222 Se Se re ee eS ee ee SS ee e e e pqjv vsodsoyoucyy e e e e pijofisuo] D1asoiq e e e e © e e © e e ee ® | ° e e e e e e e pyofyod ppawospuy e e e e e e e e e © e e e e e © e e e e e e re e e 8 e e SOIIOIKXO WINIUIIIDA wnsDdsfiSso e e e e e e e e e e © e e ® e e e ° ® e © ” r ® e e ) e LT RED ELON e e e @ e e e e e e @ e e e e ® e e ® e e e e e ® e e e pyofipunios D4sasoiq e wnsoydoyisf e e e @ e e e ® ® ° 8 e e e e e e pugdy snosayoig wnsoydoyouy @ @ @ ® © e ) @ ) @ @ e @ ® e @ ® @ @ ® @ e@ e ® @ ® pugdy snosajjosd-uoN, aim Aa][eA QIN pastel JO JOxULIG 8% LO 9 St Wve ts aCe Clow abl o> Ol ‘OU SIT satoadg ee ee a Eee ee ee ee ee eS 8 ee eee € FIAVL NI SAWIW AHL NI /YALSYTOI ‘dSANSOHLON WASOLIdSAI WNYOHdOHIINL ONIANVdAWODOOV SHIOddS “by TEV TRICHOPHORUM CESPITOSUM NOTHOSUBSP. FOERSTERI 223 Ficure 7. The distribution of 7. cespitosum subsp. cespitosum in the British Isles, based on records in 1995-1997 (@) and herbarium records (0). bogs with raised bogs, stated that the latter are of much greater age than the former; could the proliferous hybrid be a glacial relic? On the other hand, proliferation may be merely environmentally induced. For example, Lawrence (1945) showed that plants of Deschampsia cespitosa from Scandinavia, when transplanted to California, became proliferous. In Coom Rigg Moss (Mire 8), apart from the usual proliferous form with a petaloid perianth, in one area were found many plants of a proliferous form with ordinary bristles. One tuft of the proliferous plant was found on 16 July 1994 at a site quite different from any of the above. This was on a peaty, grassy slope on Carter Fell (v.c. 67) at an altitude of 550 m, at NT/687.062. This had short stomata like subsp. cespitosum, although its substomatal cavities were slightly less deep than those of the latter. This plant requires further investigation as it could be proliferous subsp. cespitosum, rather than the proliferous hybrid. DISTRIBUTION IN THE BRITISH ISLES Collections of T. cespitosum at BEL, BM, DBN, E, HAMU, K, LIV, NMW, PTH and SUN were searched and British specimens which, from visual inspection (without interfering with the specimen) appeared to be either subsp. cespitosum (Fig. 7) or the hybrid (Fig. 8) and for which it was possible to give an approximate grid reference, were mapped, along with my records (mainly from Northumberland). Among the herbarium specimens were a few of the proliferous hybrid and proliferous subsp. germanicum and these were mapped, along with the corresponding G.A.S. records (Fig. 9). Herbarium specimens of non-proliferous subsp. germanicum were also mapped for comparison (Fig. 10). Identifications of the hybrid and subsp. cespitosum were checked by cutting stem-sections, when permission to do so was granted. 224 G. A. SWAN FicureE 8. The distribution of non-proliferous Trichophorum cespitosum nothosubsp. foersteri in the British Isles, based on records by G.A.S. in 1988-1996 (@) and herbarium records (O). HERBARIUM SPECIMENS OF SUBSP. CESPITOSUM Very few specimens which were clearly subsp. cespitosum were found, and these are detailed below. Many of them are at low altitude and not in areas of the highest rainfall. v.c. 40, Salop, Twyford Vownog, near West Felton (SJ/3.2), 26 June 1840, W. A. Leighton (BM). A good, fruiting specimen. Also in E*. v.c. 58, Cheshire, Wybunbury Moss (SJ/7.4), 27 August 1892, J. E. Nowers (SUN). v.c. 83, Midlothian, Balerno Common (NT/1.6), May 1878, M. W. Evans (E)*, also June 1931, W. R. McNab (DBN). v.c. 89, East Perth, Ben Vuroch, (c. NO/0.7), 24 July 1884, J. Brebner (PTH). v.c. 112, Shetland, Bressay (HU/5.4) (NMW), with insular aerenchyma. v.c. H15, S. E. Galway, bog on shore of Lough Derg, near Woodford (R/74.99), 22 June 1898 (DBN). v.c. H18, Offaly, bog near Tullamore (N/34.25), 25 May 1895, R. L. Praeger (DBN). v.c. H38, Co. Down, Ballygowan (J/43.64), bogs, June 1903, C. H. Waddell (BEL). *These specimens in E had already been recognised and annotated as subsp. cespitosum by Dr H. A. P. Ingram in 1963. HERBARIUM SPECIMENS OF THE NON-PROLIFEROUS HYBRID v.c. 48, Merioneth, Llyn Morwynion, Ffestiniog (SH/7.4), 21 May 1938, VN. Woodhead (NMW). v.c. 57, Derbys., Goyt’s Moss, Buxton (SK/0.7), May 1883, C. T. Green (LIV). v.c. 59, S. Lancs., Ashworth Moor (SD/8.1), July 1853, Miss Graham (LIV). v.c. 64, Mid-W. Yorks., moorland on Ingleborough (SD/7.7), 31 May 1953, V. Gordon (LIV). TRICHOPHORUM CESPITOSUM NOTHOSUBSP. FOERSTERI 225 FicureE 9. The distribution of proliferous Trichophorum cespitosum nothosubsp. foersteri in the British Isles, based on records by G.A.S. in 1992-1996 (@) and herbarium records (O); and of proliferous subsp. germanicum (e), all herbarium records, except for that at NY/8.8, which is by G.A.S. in 1995. v.c. 66, Co. Durham, Widdybank Fell (NY/8.2), 14 July 1905, 7. J. Foggitt (BM). v.c. 67, S. Northumb., Fozy Moss (NY/8.7), 17 May 1957, E. F. Greenwood (LIV). v.c. 69, Westmorland, flush in blanket bog, 1 mile [1-6 km] S.W. of Moor House (NY/7.3), 1 September 1956, F. Rose (NMW). v.c. 69b, Furness, Foulshaw Moss (SD/2.7), 10 May 1913, R. S. Adamson (BM). v.c. 70, Cumberland, Butterburn Flow (NY/6.7), June 1964, F. Rose (NMW). v.c. 72, Dumfries, Lochmaben Moss (NY/0.8), R. Boyle (LIV). v.c. 83, Midlothian, Threipmuir (NT/1.6), 31 May 1834; Auchencorth Moss, near Penicuik (NT/19.55), June 1870, W. Evans (E). v.c. 87, W. Perth, Blair Drummond Moss (NS/7.9), 1 July 1882, F. B. White (PTH); bog near the col between Am Binnein and Ben More at 1700 feet [520 m] (NN/4.2), 29 July 1914, E. S. Marshall (BM, NMW). v.c. 88, Mid Perth, Stuc a Chroin, above 2800 feet [850 m] (NN/6.1), 16 July 1885, F. B. White (PTH), Breadalbane, July 1885, W. B. Waterfall (K); Ben Heasgarnich (NN/4.3), 20 July 1886, F. B. White (PTH); Rannoch Moor, near station, 950 feet [290 m] (NN/4.5), 26 June 1936, J. E. Lousley (NMW); wet moors near Ben Lawers (NN/6.4), 16 June 1946 (LIV); Ben Lawers (NN/6.4), 24 June 1950, V. Gordon (LIV); Ben Lawers, W side of N ridge, 3400 feet [1050 m] (NN/6.4), 14 July 1954, A. W. Stelfox (LIV); Craig Laoghain, N of Meall Ghaordie, with Carex saxatilis, C. nigra and Eriophorum angustifolium (NN/515.406), 700 m, 16 July 1981, A. C. Jermy, K. P. Kavanagh & A. M. Paul (BM). v.c. 89, E. Perth, Gleann Beag, above 1100 feet [335 m] (NO/1.7), 14 July 1885, A. Sturrock (PTH). 226 G. A. SWAN Figure 10. The distribution of specimens of Trichophorum cespitosum subsp. germanicum which were found in the herbaria from which the records in Figs 7, 8 and 9 were derived. v.c. 92, S. Aberdeen, end of Loch Callater (NO/1.8), 24 July 1882, F. J. Hanbury (BM); Linbrig, near Braemar (NO/1.9), 19 July 1885, F. J. Hanbury (BM). v.c. 94, Banffs, Cairngorms, Coire Raibiert, 3500 feet [1050 m] (NJ/0.0), 21 July 1966, E. Rosser (E). v.c. 96, Easterness, Glen Affric, N.E. of Loch Beneveian, among Sphagnum in wet slope in birch wood, 740 feet [225 m] (NH/2.2), 13 July 1947, E. Milne-Redhead (K); Cairngorm, peat bog, 2300 feet [700 m] (NJ/0.0?), 23 June 1964, J. K. Smith (LIV). v.c. 97, Westerness, marshes by the wood 3-4 miles [5-7 km] E. of Loch Laggan Hotel (NN/5.8), 15 July 1915, F. J. Hanbury (BM); Loch Hournhead (NG/9.0), 7 July 1949, M. S. Campbell & A. R. Clapham (BM); Glen Dubh near Alltachonaich, Morven (NM/7.5), 5 July 1973, S. S. Hooper & C. C. Townsend (K). v.c.100, Clyde Is., Beinn “a Chliabhain, Arran (NR/9.4), June 1890, A. Somerville (E). v.c. 101, Kintyre, Killean (NR/7.4), 1934, E. M. Hall (E). v.c. 105, W. Ross, small lochan to S of Loch Maree Hotel (NG/9.6), 21 July 1931, A. J. Wilmott (BM); Beinn Eighe Reserve, Allt Coir “a Laoigh, c. 400 feet [120 m], bog by road (NG/ 977.580), 30 June 1952, B. W. Ribbons, R. J. Fenn, J. T. Forrest & T. T. Mac Connell (E). v.c. 106, E. Ross, Swordale, moorland, 1100 feet [335 m] (NH/5.6), A. Meinertzhagen (BM). v.c. 108, W. Sutherland, Loch Aisa to Sandwood Loch (NC/2.6), 14 June 1949, E. B. & J. F. Basdon (BM). v.c. 110, Outer Hebrides, Lewis, R. Vowell (DBN). v.c. H18, Offaly, Tullamore (N/34.25), bog, 25 May 1895, R. L. Praeger (DBN). TRICHOPHORUM CESPITOSUM NOTHOSUBSP. FOERSTERI 227 v.c. H20, Co. Wicklow, Wicklow, R. Barrington (DBN). v.c. H25, Co. Roscommon, 1848 (DBN). v.c. H28, Co. Sligo, Knocknarea (G/50.27), June 1860, Moore (DBN); Rock Mountain, wet heath, August 1894, R. L. Praeger (DBN). v.c. H39, Co Antrim, Black Mountain, Belfast, 10 June 1878, S. A. Stewart (BEL). Also in DBN. HERBARIUM SPECIMENS OF THE PROLIFEROUS HYBRID v.c. 46, Cards., on a small flush below the forestry road, 500 m W of Bryn Mawr and 1-5 km S.E. of Hafod, with Rhynchospora alba, Carex dioica, Drosera spp. and Eriophorum angustifolium, 1100 feet [335 m] (SN/769.721), 31 August 1963, A. O. Chater (NMW). v.c. 87, W. Perth, Monachyle Glen, Balquhidder (NN/4.2), 25 August 1902, W. E. Evans (E). v.c. 97, Westerness, Glen Dubh, near Alltachonaich, Morven (NM/7.5), 5 July 1973, S. S. Hooper & C. C. Townsend (K). v.c. 107, E. Sutherland, hill N.E. of Achentoul Lodge, W. of Kinbrace station, in Erica tetralix, Calluna vulgaris, Myrica gale and Narthecium ossifragum, 550 feet [170 m] (NC/8.3), 9 July 1963, V. B. Summerhayes & P. F. Hunt (K). v.c. 108, W. Sutherland, Scourie (NC/1.4), 14 July 1885, F. J. Hanbury (BM). v.c. 110, Outer Hebrides, Lewis (NB/1.2), two sites, E. F. Warburg (BM). v.c. H33, Fermanagh, bog at Shea North (H/0.6), c. 1500 feet [460 m], 14 July 1954, R. D. Meikle (K). v.c. H39, Co. Antrim, Garron Plateau, 14 July 1979, Dr Ledsham (BEL). HERBARIUM SPECIMENS OF PROLIFEROUS SUBSP. GERMANICUM v.c. 45, Pembs., bogs on Prescelly Mountains (SN/0.3), 9 September 1932, C. I. & N. Y. Sandwith (K). v.c. 46, Cards., Rhos Rhudd Bog near Berth Rhys (SN/5.7), July 1958, G. T. Goodman (NMW). v.c. 88, Mid Perth, Killin (NN/5.3), 14 August 1918 (E). v.c. 97, Westerness, Loch Hournhead (NG/9.0), 3 July 1949, M. S. Campbell (BM). Specimen unsatisfactory. v.c. 100, Clyde Is., Gleann Easan Biorach, Arran (NR/952.475), boggy moorland, 27 August 1951, R. S. Green (E). v.c. 101, Kintyre, Killean, Kintyre (NR/7.4), 1934, E. M. Hull (E). v.c. 103, Mid Ebudes, bog between two small lochans, near Benmore Lodge, Salen, Mull (NM/556.372), 15 July 1965, H. McAllister (BM). v.c. 104, N. Ebudes, W. side of Mullach Mor by Kilmory Glen, Rhum, peat bog on sandstone (NG/3.0), 27 July 1959, A. C. Jermy (BM). v.c. 105, W. Ross, Kinlochewe (NH/0.6), 18 July 1931, and bog above hotel, 22 July 1931, A, J. Wilmott (BM); above Achnashellach, near the Clair Loch (NH/0.5) 4 August 1936, S. Sanderson (BM); Allt a’ Chiurn, c. 900 feet [275 ml], flushed marshy area (NH/003.609), 23 June 1952, A. L. C. Robertson & D. G. Moulten (E); An Teallach, c. 1000 feet [305 m], boggy ground (NH/0.8), 1974 (E).- v.c. 106, E. Ross, flush by road, Glascarnoch reservoir (NH/3.7), 21 July 1971, U. Duncan (E). v.c. 108, W. Sutherland, S. of Sandwood Loch (NC/2.6), peat bank, 9 July 1948, P. Marler (E). v.c. H35, W. Donegal, Banagher Hill, N. of Donegal town (G/9.8), rough grazing, upland, 15 July 1970 (DBN). OTHER HERBARIUM SPECIMENS The following proliferous specimen has substomatal cavities like the hybrid (14 um depth), but has a long sheath-opening, with aerenchyma islands and a petaloid perianth: v.c. 69, Westmorland, Swindale, S. side of Swindale Foot Crags (NY/517.138), 360 m, in peat bog with Eriophorum angustifolium, Erica tetralix and Sphagnum, pools with Utricularia minor, 3 August 1994, R. W. M. Corner (Herb. G.A.S.). The following non-proliferous specimen has substomatal cavities like subsp. cespitosum, but has a long sheath-opening: v.c. H18, Offaly, Seagull Bog, a few miles S. of Tullamore (N/3.2), May 1895 (DBN); see Praeger (1894). Specimens were found of non-proliferous subsp. germanicum from v.cc. H8, 18, 20, 25, 28, 29, 30, 33, 39 and 40 without locality. In addition, very many herbarium specimens are not included because the locality could not be deciphered from the label. 228 G. A. SWAN LITERATURE RECORDS Subsp. cespitosum has been recorded in the mire at Tregaron (Godwin & Conway 1939), in Meathop Moss (Ostenfeld 1912) and Flanders Moss (Professor A. Skogen, pers. comm., 1992), but in the absence of voucher specimens, it seems more likely that the plant would be the hybrid, as is the case in the Northumberland mires (M18). In other cases, such as the Silver Flow in Galloway (Ratcliffe & Walker 1958), and mires of Stainmore (Pearsall 1941), Cheshire and Shropshire and Flint (Hardy 1939), the subspecies of the Trichophorum is not stated. Moore (1968) allotted subsp. germanicum to the association in such mires, but this must surely be an error. RECENT RECORDS FROM SCOTLAND Wheeler et al. (1983) have described how in Perthshire Schoenus ferrugineus grows in base-rich flushes in a mosaic of runnels and stony hummocks. On 4 June 1997, Dr R. A. H. Smith kindly collected (at my request) two specimens of Trichophorum cespitosum from one such site in Mid Perth (v.c. 88) (B). One stem, collected from an area which was relatively less base-rich, with Erica tetralix and Myrica gale, was subsp. germanicum. The other stem, from a highly calcareous area, with Carex hostiana, C. panicea, Eriophorum latifolium and Pinguicula vulgaris appeared to be subsp. cespitosum, although final confirmation awaits finding a fruiting specimen. On 22 September 1997, Dr Smith collected a fruiting specimen from a second such Schoenus ferrugineus site in E. Perth (v.c. 89) (A) and I identified this as subsp. cespitosum. Dr R. W. M. Corner collected a specimen on 26 June 1997 from around the edge of a bog pool, which also had Drosera longifolia, Schoenus nigricans and Utricularia intermedia, on Ceathramh Garbh, W. of Rhiconich (altitude 75 m), in W. Sutherland (v.c. 108) and I identified this as subsp. cespitosum. McVean & Ratcliffe (1962) have described a site at Loch Buine Moir, Inverpolly, W. Ross (v.c. 105) which contains Trichophorum cespitosum in association with other species of Dr Corner’s site. ABSENCE OF PROLIFEROUS FORMS OUTSIDE THE BRITISH ISLES The proliferation of the spikelets (“pseudo-vivipary”) in British grasses has been studied by Wycherley (1953a, b). True vivipary is the germination of seeds while still attached to the parent plant (Raven & Walters 1956). “Viviparous” grasses such as Deschampsia cespitosa subsp. alpina, Festuca vivipara and Poa alpina may be glacial relics. According to Jermy et al. (1982), there is no record of vivipary in Carex. I found no record of proliferous T. cespitosum, with the exception of records of two sites in E. Ross (v.c. 106) (Duncan 1980), evidently referring to proliferous subsp. germanicum. According to Professor A. Skogen (Bergen) (pers. comm., 1993 and 1995), “vivipary” in T. cespitosum is very rare in Norway. He has seen it only two or three times in the field, and these were stems hanging into wet depressions so that submersion was probably the cause [i.e. possibly true vivipary, rather than floral proliferation, G.A.S.]; all belonged to subsp. germanicum. Professor H. C. Prentice (Lund) (pers. comm., 1993) stated that proliferous Trichophorum cespitosum was unknown to her and to botanical colleagues (T. Karlsson and J. T. Johansson) in Sweden. Professor H. J. B. Birks (Bergen), (pers. comm., 1995) stated that he had seen the proliferous hybrid only twice, both times in mainland Scotland - in the Cairngorms in 1967 and in Caenlochan Glen in 1976. I have been unable to find a specimen of a proliferous form from outside the British Isles, despite looking through the sheets of T. cespitosum in B, M, W and WU. Festuca vivipara requires a wet climate to enable its propagules to root, so is particularly common in N. W. Scotland. Perhaps the same is (or has been) true of the proliferous Trichophorum hybrid and especially proliferous subsp. germanicum. DISTRIBUTION IN GERMANY, HOLLAND AND BELGIUM Although subsp. germanicum is the common and widespread plant in Germany, it is not the only subspecies to occur there. Specimens exist of subsp. cespitosum from areas above c. 600 m between Munich and the Zugspitze, such as Oberammergau (BM) and near the Starnberger See and Bad Télz (M). This subspecies also occurs at low altitude in Prussia and Schleswig-Holstein in the north of Germany. There are specimens (K) from around 1900 from near K6nigsberg (now Kaliningrad, in Russia), from raised mires, 4 m above sea-level, with Eriophorum vaginatum, Rhynchospora alba, Drosera rotundifolia, Empetrum nigrum, Calluna vulgaris, Scheuchzeria palustris and Sphagnum sp. TRICHOPHORUM CESPITOSUM NOTHOSUBSP. FOERSTERI 229 However, it was Foerster’s short paper in 1963 (in which he believed he had found subsp. cespitosum growing with subsp. germanicum in the Hohes Venn) which led to Oberdorfer’s extensive work in Schwarzwald and Vosges. Paradoxically, Foerster’s supposed subsp. cespitosum from the Pannensterz proved to be the hybrid, as I had already correctly guessed (see section of the present paper, Identification, Fruiting). Nevertheless there is in M a specimen of subsp. cespitosum, which Foerster collected in 1964 in the Hohes Venn, but apparently in a different locality, and also one which he collected in Schwarzwald. Oberdorfer (1969) described various habitats for subsp. cespitosum, including flushes and raised mires and one at 1000 m altitude in the Hornisgrinde region of Schwarzwald, where subsp. germanicum and subsp. cespitosum grew side by side, although sharply separated, the latter in wet depressions in the eroded peat (like subsp. germanicum and the hybrid in Northumberland). Oberdorfer’s identification of subsp. cespitosum was evidently based almost entirely on the sheath-opening/blade length character and Fig. 1 in the present paper shows quite appreciable overlap between-subsp. cespitosum and the hybrid. He does not mention substomatal cavities or fruiting and I have found no voucher specimen. It is therefore likely that at least some of his records for subsp. cespitosum could have represented the hybrid. Through Dr W. Lippert I received from Dr. E Foerster a copy of a paper which the latter had apparently never completed. In this paper Foerster stated that in 1969 R. Tiixen had sent him specimens from moors of the Oberharz, of a plant not identifiable as either subsp. germanicum or subsp. cespitosum, using the key in Foerster (1963), so in 1970 Foerster visited the Oberharz. He found that on moors, on wet paths, and above all in the Molinia caerulea phase of the moor, subsp. germanicum occurred, but on the intact moor surface there was a population in which the stem anatomy was that of subsp. cespitosum, but in which those characters which are recognisable macroscopically were apparently intermediate between those of subsp. germanicum and subsp. cespitosum. These plants were abundant and often developed in ring-form and with a diameter sometimes well over 1 m. This indicates a very slow or quite stagnant growth of the mire over a long period (standstill complex), or an extraordinarily great age of the individual. In the following year he also found similar plants in the low country of N. W. Germany and in other German highlands. The occurrence of subsp. cespitosum has been claimed in the Netherlands by Kern et al. (1947) and Reichgelt (1956). In his paper, Foerster reported that he had investigated the voucher specimens in L on which this occurrence was based and that, according to their macroscopically recognisable characters, they belonged to the same “tribe” as the plants from Harz. He proposed to publish this as a new subspecies: Trichophorum cespitosum (L.) Hartman subsp. hercynicum subsp. nov. Foerster’s manuscript contains excellent drawings showing the sheath/blade character and stem cross-sections and gives depths of the substomatal cavities in subsp. germanicum, cespitosum and hercynicum. These all agree quite well with those from the corresponding material from Northumberland, the hybrid from the latter corresponding to subsp. hercynicum. Surprisingly, there is nothing in his manuscript to suggest that he thought subsp. hercynicum might be a hybrid of the other two subspecies, or that it failed to fruit. I propose that this hybrid be named Trichophorum cespitosum nothosubsp. foersteri, in view of the plant having been found by E. Foerster in the Harz Mountains. Although not mentioned by Foerster, the occurrence of subsp. cespitosum has been claimed by Dumont (1976) in Haute Ardenne in Belgium, who states “An unequivocal identification in the field requires only an examination of the uppermost leaf and a comparison between the lamina length and the length of the sheath opening”. However, in my opinion, this is quite insufficient and Dumont’s illustrations of stem cross-section represent the hybrid, rather than subsp. cespitosum. The description of habitat given is very similar to that of the hybrid in the Border Mires. 230 G. A. SWAN FORMAL TAXONOMY Trichophorum cespitosum (L.) Hartman Synonymy Scirpus cespitosus L. KEY TO TAXA (FOR NON-PROLIFEROUS PLANTS) L: Sheath-opening > 2 mm; length of sheath-opening/blade length > 0-4; fruit produced sed URAL yO Ee VES CAEN TE ERNE ME SOON | Yaa ROE mere ees subsp. germanicum la. Sheath-opening < 2 mm; length of sheath-opening/blade length < 0-4; fruit produced or TOE, oosssuenos ones asshs-pezne panes cuannsensents sodas sacecddne Cee enemas otucesdonsece loa natace cok es eae ae ae rrr 2 2: Sheath-opening < 1 mm (its margin without red colour); length of sheath-opening/blade length <0:25- fruit produced: 2032720) Ok Pe ee eee ete ie mena subsp. cespitosum 2a. Sheath-opening 1-2 mm (its margin possibly red-dotted); length of sheath-opening/blade length’<0:4; frurtnot produced) Wr reese ee eee nothosubsp. foersteri It seems that two taxa of Trichophorum sometimes become interwoven in the same tuft of Trichophorum and this can lead to confusion in identification. SUBSPECIES a) subsp. cespitosum Synonymy T. austriacum Palla, T. cespitosum subsp. austriacum (Palla) Hegi, S. cespitosus subsp. cespitosus, S. cespitosus var. callosus Bigelow Description: opening in upper sheath suborbicular, 1 mm in diameter, with blade 5-10 x as long as opening (length of sheath-opening/blade length < 0-25); plant fruiting; aerenchyma absent; substomatal cavities 20-26 um deep. Distribution: Arctic-Alpine b) subsp. germanicum (Palla) Hegi Synonymy T. germanicum Palla, S. cespitosus subsp. germanicus (Palla) Broddeson, S. germanicus (Palla) Lindman Description: opening in upper sheath oblique, 2—3(—5) mm long and | mm broad, with blade up to 2 x as long as opening (length of sheath-opening/blade length > 0-4); plant fruiting; aerenchyma present; substomatal cavities 6—7 um deep. Distribution: Atlantic-Subatlantic Trichophorum cespitosum (L.) Hartman subsp. cespitosum x subsp. germanicum (Palla) Hegi = nothosubsp. foersteri G. A. Swan, nothosubsp. nov. Per rationem foraminis vaginae contra longitudinem laminae foliae 0-2—0-35, cavernulae stomatis profunditate 8-18 um, fructu sterili parentibus subsp. cespitoso (ratio <0-25, cavernulae profunditas 20-26 um, fertilis) et subsp. germanico (ratio > 0-4, cavernulae profunditas 6—7 um, fertilis) differt. Differing in the ratio of length of sheath-opening to length of blade 0-2—0-35, the depth of the substomatal cavities 8-18 um, and sterile fruit, which in subsp. cespitosum are ratio < 0-25, cavities 20-26 um deep, fertile and in subsp. germanicum are ratio > 0-4, cavities 6-7 um deep, fertile. Hotortypus: S. Northumberland, v.c. 67, Car Knowe Moss (NY/727.787), raised mire, 20 July 1996, G. A. Swan (BM). TRICHOPHORUM CESPITOSUM NOTHOSUBSP. FOERSTERI 231 SUMMARY The hybrid is the predominant plant of raised mires, not only in Northumberland, but also in the Harz Mountains in Germany; it occurs also in S. W. Germany, Holland, Belgium and Norway, as well as elsewhere in Britain and Ireland. In earlier times, subsp. cespitosum presumably covered areas which the hybrid does today. The survival of the sterile hybrid is reminiscent of that of Circaea x intermedia (Raven 1963) and Nuphar x spenneriana (Heslop-Harrison 1953). Many records for subsp. cespitosum represent this hybrid, so populations of T. cespitosum in these areas should be re-examined. A plant which has a sheath-opening c. 1 mm and which is producing mature fruits in short, broad heads is likely to be subsp. cespitosum. A plant with a sheath-opening 1-1-5 mm and in which the top of the stem becomes more or less bare by July is likely to be the hybrid. In either case, the identification should be confirmed by microscopic examination of the stem-section. Although subsp. cespitosum does not in general require base-rich conditions, in regions where subsp. germanicum is also present, subsp. cespitosum may be found only in base-rich areas because only there has it escaped hybridisation; attempts should be made to find out whether or. not this statement is generally true. ACKNOWLEDGMENTS Thanks are offered to the curators of B, BEL, BM, DBN, E, HAMU, K, LIV, M, NMW, PTH, SUN, W and WU for access to and loan of material and to A. Coles for facilitating loans to the Hancock Museum, Newcastle-upon-Tyne. Thanks are also due to Gina Douglas for access to the Linnaean Herbarium in London, to the Botany Library of the Natural History Museum, London and the University Library, University of Newcastle-upon-Tyne for facilitating access to literature. Thanks are offered to Mrs O. M. Stewart for drawing Fig. 6, to C. D. Preston for preparing the distribution maps, to D. H. Kent and N. J. Turland for advice on nomenclature, to Mrs M. Patterson for typing the manuscript, and to A. C. Jermy, A. O. Chater and Dr A. G. Lunn, with whom I had discussions at an early stage of the work. It was Mr T. and Mrs D. Hardy’s excitement in finding Menyanthes trifoliata flowering (unusual at 300 m) at Blackheugh End that caused me to visit that site on 17 July 1995 and so find T. cespitosum subsp. cespitosum for the first time in Northumberland; I thank them. I would also like to thank Dr E. Foerster for specimens and a copy of his manuscript, and Dr W. Lippert for facilitating this. Also my wife, Margaret, who accompanied me in visiting almost all of the mires and helped with translation from Norwegian. Finally, Dr A. J. Richards kindly took the microphotographs, provided the Latin diagnosis and was always prepared to give advice, and Dr P. M. Hollingsworth who kindly carried out the isozyme work in support of my conclusion. REFERENCES BEETLE, A. A. (1947). North American Flora 18 (8): 494. BIGELow, J. (1824). A collection of plants of Boston and its vicinity, 2nd ed. Cummings, Hillard and Co., Boston. Bropbeson, E. (1912). Om de skandinaviska formerna af Scirpus caespitosus. Botaniska notiser 1912: 81-94. CLAPHAM, A. R., TuTIN, T. G. & Moore, D. M. (1987). Flora of the British Isles, 3rd ed. Cambridge University Press, Cambridge. CLAPHAM, A. R., TuTIn, T. G. & WARBUuRG, E. F. (1952). Flora of the British Isles. Cambridge University Press, Cambridge. CLAPHAM, A. R., TuTIn, T. G. & WarBurG, E. F. (1962). Flora of the British Isles, 2nd ed. Cambridge University Press, Cambridge. CLAPHAM, A. R., TuTin, T. G. & Warsura, E. F. (1981). Excursion Flora of the British Isles, 3rd ed. Cambridge University Press, Cambridge. De Fiuipps, R. A. (1980). Scirpus L., in Tutin, T. G. et al., eds. Flora Europaea 5: 277-280. Cambridge University Press, Cambridge. Dumont, J.-M. (1976). Les deux sous-espéces de Scirpus cespitosus L. en Haute Ardenne, particuliérement au Plateau des Tailles. Bulletin de la Société Royale de Botanique de Belgique 109: 307-318. 2352 G. A. SWAN Duncan, U. K. (1980). Flora of East Ross-shire. Botanical Society of Edinburgh, Edinburgh. ELLENBERG, J. (1988). Vegetation ecology of Central Europe, 4th ed. English translation. Cambridge University Press, Cambridge. FERNALD, M. L. (1921). The North American representatives of Scirpus cespitosus. Rhodora 23: 22-25. Firsas, F. (1931). Untersuchungen tiber den Wasserhaushalt der Hochmoorpflanzen. Jahrbuch fiir wissenschaftliche Botanik 74: 459-696. Foerster, E. (1963). Trichophorum caespitosum (L.) Hartm. ssp. caespitosum im Hohen Venn. Decheniana 115: 274-275. FREMSTAD, E. & SKOGEN, A. (1978). Trichophorum caespitosum ssp. germanicum in Norway. Blyttia 36: 135-144. Gopwin, H. & Conway, V. M. (1939). The ecology of a raised bog near Tregaron, Cardiganshire. Journal of ecology 27: 313-363. Harpy, E. M. (1939). Studies of the post-glacial history of British vegetation. V. The Shropshire and Flint Maelor Mosses. New phytologist 38: 364-396. HEaI, G. (1909). I/lustrierte Flora von Mitteleuropa, \st ed., 2: 24-26. Lehmann’s Verlag, Miinchen. HEI, G. (1966). Illustrierte Flora von Mitteleuropa, 3rd ed., 2 (1): 37-39. Carl Hanser Verlag, Miinchen. HEsLop-Harrison, Y. (1953). Nuphar intermedia Ledeb., a presumed hybrid in Britain. Watsonia 3: 7-25. Hess, H. E., LANDOLT, E. & HirzeEL, R. (1967). Flora der Schweiz und angrenzender Gebiete 1: 400-402. Birkhauser, Basel and Stuttgart. HOLLinGsworTy, P. M. & Swan, G. A. (1999). Genetic differentiation and hybridisation among subspecies of Deergrass (Trichophorum cespitosum (L.) Hartman) in Northumberland. Watsonia 22: 235-242. HuLtTEn, E. (1962). The circumpolar plants. I. Vascular cryptogams, conifers, monocotyledons. Kungliga Svenska vetenskapsakademiens handlingar, Series 4, 8 (5): 1-275. JermMyY, A. C., CHATER, A. O. & Davin, R. W. (1982). Sedges of the British Isles, 2nd ed. B.S.B.I. Handbook No. 1. B.S.B.I., London. KENT, D. H. (1992). List of vascular plants of the British Isles. B.S.B.I., London. KERN, J., REICHGELT, B. & REICHGELT, T. (1947). De Ondersoorten van Trichophorum caespitosum (L.) eantir in Nederland. Nederlandsch kruidkundig archief 54: 260-263. Koyama, T. (1958). Taxonomic studies of the genus Scirpus. Journal of the Faculty of Science, Tokyo University, Section 3, Bot. 7: 353-354. LAwRENCE, W. E. (1945). Some ecotypic relations of Deschampsia caespitosa. American journal of botany 32: 298-314. Lunn, A. G. & Lunn, J. (1976). Nature reserves of the Northumberland Wildlife Trust. The Northumberland Wildlife Trust Ltd, Newcastle-upon-Tyne. Lussy, P. & WRIGHT, J. (1996). Scottish wild plants. Edinburgh: The Stationery Office, London. McVEAan, D. N. & RatcuiFFE, D. A. (1962). Plant communities of the Scottish Highlands. H.M.S.O., London. METSAVAINIO, K. (1931). Untersuchungen iiber das Wurzelsystem der Moorpflanzen. Annales botanici Societatis zoologicae - botanica fennicae Vanamo. Tome 1, No. 1. MEusEL, H., JAGER, E. & WEINERT, E. (1965). Vergleichende Chorologie des zentral-europdischen Flora. Karten, 59. Jena. MontrorT, C. (1918). Die Xeromorphie der Hochmoorpflanzen als Voraussetzung der “physiologischen Trockenheit” der Hochmoor. Zeitschrift fiir Botanik 10: 257-352. Moore, J. J. (1968). A classification of the bogs and wet heaths of Northern Europe, in TUxeEN, R. ed. Pflanzensoziologisce Systematik, pp. 306-320. Den Haag: Junk, N.V. OBERDORFER, E. (1969). Zur Verbreitung und Soziologie von Trichophorum cespitosum (L.) Hartm. subsp. cespitosum und subsp. germanicum (Palla) Hegi. Berichte der Deutschen botanischen Gesellschaft 82: 589-594. OsTENFELD, C. H. (1912). New phytologist 11: 125-126. OsTENFELD, C. H. & GrontveD, J. (1934). The flora of Iceland and the Faeroes. Levin and Munksgaard, Copenhagen; Williams and Northgate, London. OsvaLp, H. (1923). Die Vegetation des Hochmoores Komosse. Svenska Vdxtsociologiska Sdllskapets Handlingar 1: 1-436. PaLLa, A. (1897). Einige Bemerkungen tiber Trichophorum atrichum und caespitosum. Berichte der Deutschen botanischen Gesellschaft 15: 467-471. PEARSALL, W. H. (1941). The “mosses” of the Stainmore district. Journal of ecology 29: 161-175. Po.unin, N. (1959). Circumpolar Arctic Flora. The Clarendon Press, Oxford. PRAEGER, R. L. (1894). The Seagull Bog, Tullamore. The Irish naturalist 3: 173-175. RATCLIFFE, D., ed. (1977). A nature conservation review. Cambridge University Press, Cambridge. RatcuirFe, D. A. & WALKER, D. (1958). The Silver Flowe, Galloway, Scotland. Journal of ecology 46: 407-445. RAVEN, J. & WALTERS, M. (1956). Mountain flowers. Collins, London. Raven, P. H. (1963). Circaea in the British Isles. Watsonia 5: 262-272. TRICHOPHORUM CESPITOSUM NOTHOSUBSP. FOERSTERI 233 REICHGELT, T. J. (1956). Flora Neerlandica 1 (4): Cyperaceae excl. Carex. Koninklijke Nederlandse Botanische Vereniging, Amsterdam. Ropwe.L, J. S., ed. (1991). British plant communities, 2 Mires and Heaths. Cambridge University Press, Cambridge. Rose, F. (1953). A survey of the ecology of British lowland bogs. Proceedings of the Linnean Society of London 164: 186-211. Rotn, W. A. (1789). Tentamen Florae germanicae 2: 53. Lipsiae. SALMENKALLIO, M. & KUKKONEN, I. (1989). Proposal to conserve 466a Trichophorum (Cyperaceae) with a conserved type. Taxon 38: 313-316. Scoacaan, H. J. (1978). The Flora of Canada, part 2: 448. National Museums of Canada, Ottawa. SELL, P. D. & Murrett, G. (1996). Flora of Great Britain and Ireland 5. Cambridge University Press, Cambridge. Syors, H. (1950). Regional studies in North Swedish mire vegetation. Botaniska notiser 1950: 173-222. SoRTLAND, A. (1992). Ny nordgrense for kystbjgnnskjegg (Trichophorum cespitosum subsp. germanicum). Polarflokken 16: 15-18. Stace, C. A. (1991). New Flora of the British Isles. Cambridge University Press, Cambridge. Stace, C. A. (1997). New Flora of the British Isles, 2nd ed. Cambridge University Press, Cambridge. Swan, G. A. (1993). Flora of Northumberland. Natural History Society of Northumbria, Newcastle-upon- Tyne. WESTERMAIER, M. (1881). Beitrage zur Kenntniss des mechanischen Gewebesystems. Monatsbericht der Koniglichen preussischen Akademie der Wissenchaften zu Berlin: 61-78. WHEELER, B. D., Brookes, B. S. & Situ, R. A. H. (1983). An ecological study of Schoenus ferrugineus L. in Scotland. Watsonia 14: 249-256. WYCHERLEY, P. R. (1953a). Proliferation of spikelets in British grasses. I. The taxonomy of the various races. Watsonia 3: 41-56. WYCcHERLEY, P. R. (1953b). The distribution of the viviparous grasses in Great Britain. Journal of ecology 41: 275-288. (Accepted June 1998) ie x. = li eS oe = aah Aine eh hg ¥ Tae Vineet Cs ae? 4 au at shaun ceng ender tava ni ee yeh sistas A lt weenie “ibis ie cia 4 Paik va a ae He. ee oe LE a i fe adivalk vil dm) moneda we wyieanon nt stony (eae ; ai ceranerarry oo var anr denne yt Aes: ever epee. bane’ te want ieeciant Ort 2 nay obanied 5 THAD nubs 3, She hae aan, ee OND TPR CE ( «Se heey fi (7 iniioew whiten at subeac te ik sot rraliireb dation Lite Sehayapirior by dh pany» nie seine etpaaiwcie me ; com er pvt a ee ad enon OY si nid sae me € Py thhiert sot) egberdn spree ent ule ak ie Lridie BL bowers abe! rey trot My IS Praoketet: diaaiatlal Monatussinenty vay api ol ce ee n hea ad atte ashiFin ct hava! its ilnenet ye A rusrtbyjeiate Whos As LAD CRAPS oe sear ebael Renae Seated e ee snndin eels Manend Up areeiast gga TAA acoipri = ad snare aa Es: ie eater verniaa uah dienes ae O2t--BAS tee Sane ra eek dan tqngetiondxa 0b ee, wna oi canton be. ouhateitrl LAE Sheva hlnss ed Va wh inidtixte wot hows ® ih th oH mn ju thapsttilT rv H a E iii ; , 70% ia 7 pug ren ay i Ys ee ee Ae URES d. Subsp. cespitosum © Margin of Gowany Knowe m 67 NY/727.788 j Blackheugh End k 67 NY/827.915 TABLE 2. ISOZYME VARIATION IN EIGHT ENZYME SYSTEMS IN TRICHOPHORUM CESPITOSUM (L.) HARTMAN Uppercase letters correspond to zymograms in Fig. 1. Population codes refer to Table 1. G6 a. Subsp. germanicum a b. Nothosubsp. foersteri (non-proliferous) c. Nothosubsp. foersteri (proliferous) uu 5 d. Subsp. cespitosum WN Q [o) W Drrn wr FPWrrr FoRVrPKP PrP? yl Ar LS ANNAN BOW www wD Prrrrrrrryvl moe ANNA BW WDWWDWD Frrrrrrry| ro ANNAN BW DOWWWTwD Frrrrrrry UA AWDW>FO DW WDWWDWDWD Frrrrrrrry| ov >>r>r>r Fr rPrPrry>r FPrPrPrrrrrr sav >rPrrr Fr rrr rr FFF rKrPr rrr TVAY >Prr>r Fr Frere FFF PrPrer eS VV AANA Wh NWN N WH Aw Ss 8 MLG = multi-locus genotype. Sample size Uo —= WN — = nan NK | = = Dn WN Sample size MNnNne nN — KS WwW = WN — — WWNnN WN Ww WH 238 P. M. HOLLINGSWORTH AND G. A. SWAN FicurE 1. Zymogram representing banding patterns in Trichophorum cespitosum (L.) Hartm. in Northumberland (anode towards the top of the figure). Letters below the banding patterns represent a single enzyme genotype, the distribution of these among samples are given in Table 2. Ger = subsp. germanicum, Foe = nothosubsp. foersteri and Ces= subsp. cespitosum. RESULTS ENZYME PHENOTYPES From the eight enzyme systems examined, a minimum of eleven anodally migrating loci was clearly resolved (Fig. 1 and Table 2). These are described in turn below. AAT - a single putative locus was resolved, although there was a faint trace of a second faster migrating locus on the gel. At the locus that was clearly resolved, all samples of subsp. germanicum were homozygous for a fast allele, nothosubsp. foersteri was either homozygous for this fast allele, or heterozygous for this and a slower allele. Subsp. cespitosum showed all possible combinations of the two alleles. G6PDH - two putative loci were resolved with both being homozygous and uniform in all three taxa. IDH - two putative loci were resolved with the most anodally migrating locus being uniform across all samples. The other locus was variable, although resolution was poor. All individuals of subsp. germanicum have a fast moving zone of activity, all individuals of subsp. cespitosum have a slow moving zone of activity, and all individuals of nothosubsp. foersteri have a smear which corresponds to both zones of activity. SUBSPECIES OF TRICHOPHORUM CESPITOSUM 239 Figure 2. MDH evidence for hybridisation between Trichophorum cespitosum subsp. cespitosum and subsp. germanicum in Northumberland (anode towards the top of the figure). Two loci are shown with all taxa being homozygous and uniform for the least anodally migrating locus. At the most anodally migrating locus, all individuals of subsp. germanicum (from left to right, lanes 1-12) are homozygous and uniform for a fast moving allele (marked A), all individuals of subsp. cespitosum (lanes 26-35) are homozygous and uniform for a slow moving allele (marked B), and all individuals of nothosubsp. foersteri (lanes 13-25) show a heterozygous genotype (including a heterodimer) for these two alleles. Bands present just behind these two different alleles are considered to be artifactual and do not represent additional alleles or loci. MDH - two putative loci were resolved with all taxa being homozygous and uniform for the least anodally migrating locus. At the most anodally migrating locus, all samples of subsp. cespitosum were homozygous and uniform for a slow moving allele, all individuals of subsp. germanicum were homozygous and uniform for a fast moving allele, and all samples of nothosubsp. foersteri showed a heterozygous genotype for these two alleles (Fig. 2). PGD - two putative loci were detected, the most anodally migrating was invariant across all samples. At the second locus, all samples of subsp. cespitosum were homozygous for a fast moving allele, all samples of subsp. germanicum were homozygous for a slow moving allele and all samples of nothosubsp. foersteri were heterozygous for these two alleles. PGI - two putative loci were detected with the most anodally migrating locus being poorly resolved but apparently uniform. At the least anodally migrating locus, all samples of subsp. germanicum were homozygous and uniform for a slow moving allele, all samples of nothosubsp. foersteri were heterozygous for this and a fast moving allele. Subsp. cespitosum showed all possible combinations of the two alleles. PGM - all taxa were monomorphic for two bands. As PGM is a monomeric enzyme it is not clear whether this represents all samples being homozygous and uniform for two separate loci, or fixed heterozygosity at a single locus. SKD - One putative locus was detected at which all taxa were homozygous and uniform. VARIATION BETWEEN TAXA Subsp. cespitosum and subsp. germanicum in Northumberland are clearly distinguishable by fixed genetic differences for three enzyme systems MDH, PGD and IDH. For all of these enzyme systems, nothosubsp. foersteri shows apparent direct additive inheritance of these taxon-specific markers. VARIATION WITHIN TAXA All individuals of subsp. germanicum showed the same multi-locus enzyme genotype, and at the loci where allelic designations could be made, all samples were homozygous. Subsp. cespitosum showed diallelic variation for two enzyme systems (AAT and PGI). A total of four different genotypes was identified, two from each of the two populations sampled (Fig. 1, Table 2). For these variable loci, both homozygous and heterozygous genotypes were recovered, although the small sample sizes and our cytological ignorance preclude tests for deviations from random mating based on the Hardy-Weinberg principle. A total of two different genotypes of nothosubsp. foersteri was detected based on variation at one enzyme system (AAT). No intra-population variation was detected, with four of the populations having one of the genotypes, and the other genotype being confined to the fifth population (Fig. 1, Table 2). It should be borne in mind, however, that for two populations, only one individual plant 240 P. M. HOLLINGSWORTH AND G. A. SWAN was sampled and the sample sizes for the other populations are small. No genetic markers were found that distinguished between the proliferous and non-proliferous plants. DISCUSSION TAXON DIFFERENTIATION AND HYBRIDISATION The sample sizes used in this study are low and we stress that the following conclusions remain tentative. Nevertheless, the initial findings are quite striking. Isozyme evidence suggests there is a clear genetic differentiation of subsp. germanicum and subsp. cespitosum in Northumberland, with three loci (MDH, IDH, PGD) showing apparently fixed genetic differences. The data also provide strong support for nothosubsp. foersteri being of hybrid origin from the two subspecies (Figs 1 & 2). The absolute concordance of data from these loci between populations (including comparisons with Scottish populations) suggests that these samples are representative for British plants, although of course further verification including wider geographic sampling is desirable. Given the fixed genetic differences between the subspecies (based on this sample) and the sterility of the hybrid, this raises the question as to whether the subspecies should be raised to specific rank. We cannot commit ourselves to this based on such a small sample, but point out that if more intensive studies support the conclusions from this work, namely a. that there are clear fixed differences between the two sub-species, and b. that the hybrids are sterile, then specific rank for subsp. germanicum and subsp. cespitosum should be considered. However, it is worth stressing that the question of the sterility of nothosubsp. foersteri remains open for discussion. Although no mature fruits have ever been seen on plants of nothosubsp. foersteri, Swan (1999) noted that there is often high stainability of its pollen; he also reported the presence of Trichophorum plants growing in Northumberland that appear to be intermediate between nothosubsp. foersteri and subsp. germanicum. These may represent backcrosses and there is an obvious need to study these plants to determine whether their isozyme profiles support this suggestion. Any possibility of trace fertility in nothosubsp. foersteri potentially complicates interpretations of the number of origins of the hybrid. AAT variation between plants of nothosubsp. foersteri could indicate that the hybrid has arisen on at least two occasions, i.e. it is the product of more than one successful gamete fusion (Fig. 1). We feel that this, rather than hybrid fertility, is the most likely explanation for the genetic variation in the samples of the hybrid examined in this study, as (a.) no disruption of the taxon specific markers (MDH, IDH, PGD) was detected amongst the hybrids, and (b.) the variation for AAT is exactly as would be predicted based on crosses between the fixed fast band of subsp. germanicum and both the fast and the slow alleles from subsp. cespitosum. In this respect it is noteworthy that the homozygous slow condition is not detected in nothosubsp. foersteri due to the invariant fast allelic constitution of subsp. germanicum. Variation of PGI in subsp. cespitosum indicates that the nothosubsp. foersteri plants in this study result from a cross involving a subsp. cespitosum plant or plants with the fast PGI allele, as no individuals of nothosubsp. foersteri homozygous for the slow allele were observed. Our inability to distinguish between the proliferous and the non-proliferous forms of nothosubsp. foersteri using isozymes may simply be a result of not sampling enough loci, and these different forms may represent additional origins of the hybrid. The difference in spikelet morphology between the two forms is quite striking, with the bristles in the non-proliferous plants being replaced by a membranous perianth in the proliferous plants. However, environmentally induced variation in spikelet morphology and function is far from unknown, and the transition from non-prolifery to prolifery is well documented (e.g. as in Deschampsia cespitosa (L.) P. Beauv.; Briggs & Walters 1984). The proliferous plants sampled here showed identical multi-locus genotypes with their sympatric non-proliferous counterparts (Table 2). In the absence of informa- tion from more molecular markers, and given our lack of understanding of what the mechanism would be if prolifery was under genetic control, it is difficult to comment further. VARIATION WITHIN TAXA The uniformly homozygote genotype detected within subsp. germanicum precludes us from distinguishing between clonal growth and sexual reproduction (be it self-pollination, mixed mating or outcrossing) as the major reproductive mode. The presence of hybrids with subsp. cespitosum, however, indicate that reproduction is probably not purely asexual or purely autogamous. SUBSPECIES OF TRICHOPHORUM CESPITOSUM 241 Large-scale sexual reproduction by nothosubsp. foersteri is considered unlikely based on the results of this isozyme survey and the fact that mature fruits have never been found (as discussed above). A more probable explanation is that nothosubsp. foersteri persists via a combination of clonal propagation and recurrent origins from the parental subspecies. Further sampling is required, however, to substantiate this. Diallelic variation and the presence of both homozygote and heterozygotes for AAT and PGI in subsp. cespitosum provides good evidence for sexual reproduction with at least some outcrossing in this taxon. That more variation was detected in subsp. cespitosum, which is local and rare in Britain, than in the common and widespread subsp. germanicum is surprising and we as yet have no explanation for this apparent paradox. A greater sampling of both individuals and loci, set in the context of a firm cytological framework, may offer further insights into the reproductive biology and population biology of these taxa. ACKNOWLEDGMENTS We are grateful to A. J. Richards for many helpful comments and advice on this project and to Chris Preston for comments on an earlier draft of this manuscript. PMH acknowledges the facilities of the University of Glasgow’s N.E.R.C. Taxonomy Initiative and the Royal Botanic Garden, Edinburgh. REFERENCES ABBOTT, R. J. (1992). Plant invasions, interspecific hybridization and the evolution of new plant taxa. Trends in ecology and evolution 7: 401-405. ARNOLD, M. L., BUCKNER, C. M. & Rosinson, J. J. (1991). Pollen mediated introgression and hybrid speciation - in Louisiana irises. Proceedings of the National Academy of Sciences, USA 88: 1398-1402. AsuTon, P. A. & AsBsotrt, R. J. (1992). Multiple origins and genetic diversity in the newly arisen allopolyploid species, Senecio cambrensis Rosser (Compositae). Heredity 68: 25-32. Briccs, D. & Watters, S. M. (1984). Plant variation and evolution, 2nd ed. Cambridge University Press, Cambridge. CLaPHaM, A. R., TuTin, T. G. & Moore, D. M. (1987). Flora of the British Isles, 3rd ed. Cambridge University Press, Cambridge. De Fiuipps, R. A. (1980). Scirpus L., in Tutin, T. G., HEywoop, V. H., Burces, N. A., Moore, D. M., VALENTINE, D. H., Watters, S. M. & Wess, D. A. eds. Flora Europaea 5: 277-280. Cambridge University Press, Cambridge. GALLEz, G. P. & Gortruies, L. D. (1982). Genetic evidence for the hybrid origin of the diploid plant Stephanomeria diegensis. Evolution 36: 1158-1167. Gortuies, L. D. (1981). Electrophoretic evidence and plant populations. Progress in phytochemistry 7: 1-46. Gort.ies, L. D. (1982). Conservation and duplication of isozymes in plants. Science 216: 373-380. HecI, G. (1909). I/lustrierte Flora von Mitteleuropa, 1st ed. 2: 24-26. Lehmann’s Verlag, Munich. HOLLINGSworTH, P. M., GorNALL, R. J. & Preston, C. D. (1995a). Genetic variability in British populations of Potamogeton coloratus (Potamogetonaceae). Plant systematics and evolution 197: 71-85. HOLLINGSworTH, P. M., Preston, C. D. & GorNALL, R. J. (1995b). Isozyme evidence for hybridisation between Potamogeton natans and P. nodosus (Potamogetonaceae) in Britain. Botanical journal of the Linnean Society 117: 59-69. HOLLINGsworTH, P. M., Preston, C. D. & GorNALL, R. J. (1996). Isozyme evidence for the parentage and multiple origins of Potamogeton xX suecicus (P. pectinatus x P. filiformis, Potamogetonaceae). Plant systematics and evolution 202: 219-232. HuLTENn, E. (1962). The circumpolar plants. 1. Vascular cryptogams, conifers, monocotyledons. Kungliga Svenska Vetenskapsakademiens Handlingar Series 4, 8(5): 1-275. JORGENSON, C. A., SORENSEN, T. & WESTERGAARD, M. (1958). The flowering plants of Greenland. A taxonomi- cal and cytological survey. Biologiske meddelelser. Kongelige Danske videnskabernes selskab 9(4): 1-172. Liston, A., RIESEBERG, L. H. & Mistretra, O. (1990). Ribosomal DNA evidence for hybridisation between island endemic species of Lotus. Biochemical systematics and ecology 18: 239-244. Love, A. & Love, D. (1956). Cytotaxonomical conspectus of the Icelandic flora. Acta Horti Gotoburgensis 20: 65-291. MEuseEL, H., JAGER, E. & WEINERT, E. (1965). Vergleichende Chorologie des Zentral-Europdischen Flora. Karten, 59. Jena. 242 P. M. HOLLINGSWORTH AND G. A. SWAN Newsury, H. J. & Forp-Lioyp, B. V. (1993). The use of RAPD for assessing variation in plants. Plant growth regulation 12: 43-51. RaYBOULD, A. F., Gray, A. J., LAwRENCE, M. J. & MARSHALL, D. F. (1991a). The evolution of Spartina anglica C. E. Hubbard (Gramineae): origin and genetic variability. Biological journal of the Linnean Society 43: 111-126. RAYBOULD, A. F., Gray, A. J., LAWRENCE, M. J. & MARSHALL, D. F. (1991b). The evolution of Spartina anglica C. E. Hubbard (Gramineae): genetic variation and status of the parental species in Britain. Biological journal of the Linnean Society 44: 369-380. RIESEBERG, L. H. & ELLsTRAND, N. C. (1993). What can molecular and morphological markers tell us about plant hybridisation? Critical reviews in plant sciences 12: 213-241. Roose, M. L. & Gortuizs, L. D. (1976). Genetic and biochemical consequences of polyploidy in Tragopogon. Evolution 30: 818-830. SCHEERER, H. (1940). Chromosomenzahlen aus der Schleswig-Holsteinischen Flora II. Planta 30: 716-725. SELL, P. D. & Murre.t, G. (1996). Flora of Great Britain and Ireland, vol. 5 Butomaceae — Orchidaceae. Cambridge University Press, Cambridge. StAcE, C. A. (1991). New Flora of the British Isles, 1st ed. Cambridge University Press, Cambridge. Stace, C. A. (1997). New Flora of the British Isles, 2nd ed. Cambridge University Press, Cambridge. Swan, G. A. (1999). Identification, distribution and a new nothosubspecies of Trichophorum cespitosum (L.) Hartman (Cyperaceae) in the British Isles and N. W. Europe. Watsonia 22: 209-233. Wana, X.-R. & Szipt, A. E. (1990). Evolutionary analysis of Pinus densata, a putative Tertiary hybrid. 2. A study using species-specific chloroplast DNA markers. Theoretical and applied genetics 80: 641-647. WEEDEN, N. F. & WENDEL, J. F. (1989). Genetics of plant isozymes, in Soxtis, D. E. & So tis, P. S. eds. Isozymes in plant biology, pp. 46-72. Dioscorides Press, Portland, Oregon. (Accepted May 1998) Watsonia 22: 243-250 (1999) 243 Towards a simplified taxonomy of Capsella bursa-pastoris (L.) Medik. (Brassicaceae) A. AKSOY’, W. H. G. HALE” and J. M. DIXON Dept of Environmental Science, University of Bradford, Bradford, West Yorkshire, BD7 1DP ABSTRACT Capsella bursa-pastoris (L.) Medik. is a species with a cosmopolitan distribution which shows considerable morphological variation. Numerous authors have recognised widely differing numbers of varieties, mi- crospecies or other infraspecific subdivisions (segregates) of this species. In an attempt to clarify this situation, we grew British material of the species under controlled conditions through to the F, generation to remove environmental variation, and assessed the plants on the basis of a range of morphological criteria, namely leaf shape, capsule size and also length of time taken to flower. Analysis of these characteristics consistently produced four basic groups, which had been previously described. Herbarium specimens could also nearly always be assigned to one of these groups. Limited chromosome counts suggest that two of these groups are diploid and two are tetraploid. We suggest this fourfold division into broad groups reflects the major genetic separations within the species, but that there is also considerable phenotypic plasticity shown by C. bursa- pastoris in response to factors such as shade or trampling. These four groups appear to differ in their geographical distribution in Britain. Keyworps: Shepherd’s Purse, morphological variation, leaf characters, capsule characters, chromosome counts, infraspecific groupings. INTRODUCTION Capsella bursa-pastoris (L.) Medik. (Shepherd’s Purse) (Brassicaceae) has a cosmopolitan distri- bution, and is a colonising species of disturbed ground. Being found in a broad range of conditions, up to 5900 m (Wilson 1949; Mani 1978) and in almost all countries of the world from tropical to subarctic habitats (Holm et al. 1979), the species is known to exhibit considerable morphological variation. Capsella bursa-pastoris has been described by numerous authors since the late 19th century, and has been divided taxonomically into many species, subspecies, varieties, microspecies and segre- gates. Jordan (1864), one of the earliest workers, described five species in France, namely Capsella agrestis, C. virgata, C. ruderalis, C. sabulosa and C. praecox, none of which are recognised today. Hopkirk (1869) considered the variation in Belgium to consist of subspecies derived from one common type, and he went on to describe six subspecies based primarily on the character of the capsule. Mott (1885) described eight varieties for Leicestershire and Rouy & Foucaud (1893) listed seven varieties and four subspecies based on the fruit characteristics in France. Almquist (1907) described 70 elementary species and later (Almquist 1921) examined British Capsella bursa- pastoris and listed 16 species. His descriptions were based on fine distinctions of leaf and capsule shape and size. Two years later, Almquist (1923) had recorded twelve classes of Capsella containing almost 200 microspecies. His microspecies were again based on minute differences in capsule shape and size, differing leaf shapes and position of leaf lobes. More recently, but only in Cyprus, Meikle (1977) recognized two species based on capsule size whilst Clapham et al. (1987) record Capsella bursa-pastoris as “very variable with a strong tendency for distinctive populations to arise because of self-pollination. Many of these have been named”, but they do not specify any of these. The first edition of Flora Europaea (Chater 1964) comments that numerous variants have been described by Almquist, whilst the second edition (Chater 1993) states that “there is extreme * Present address: Department of Biology, Faculty of Science and Art, University of Erciyes, Kayseri, Turkey. ** Corresponding author 244 A. AKSOY, W. H. G. HALE AND J. M. DIXON polymorphism within the four species listed” and that “Capsella bursa-pastoris is especially polymorphic and its variants incorporate many of the characteristics of the other three species”. Most recently, Stace (1997) describes C. bursa-pastoris “‘as extremely variable in leaf and fruit shape; c. 25 segregates have been recognized in the British Isles”. No details of these are given, nor iS an information on them provided by the specialist Crucifers of Great Britain and Ireland (Rich 1991). The problems of taxonomy at the infraspecific level, relevant to a very variable species such as C. bursa-pastoris which is largely inbreeding yet has many phenotypic variations, are discussed, for example, by Stace (1989). This author notes that such phenotypic modifications would not be given taxonomic status by most taxonomists and that when such variations are recognised as phenotypic, they are relegated to synonomy. Capsella bursa-pastoris certainly shows phenotypic variation as a direct result of a wide range of environmental factors including temperature, shading, altitude, latitude and season; for example, Almquist (1923) found that leaves developing in autumn and spring were mostly lobed, whereas mid-summer leaves tend to be poorly lobed or entire. Hurka (1990) found pronounced ecotypic variation in time to flowering between early Scandinavian and late Alpine populations; he also found early and late ecotypes in North America. Aksoy (1996) observed the effects of shading on leaf shape, Steinmayer et al. (1985) recorded correlations between leaf form and temperature and rainfall, and Neuffer (1989) investigated the effects of temperature on variables such as leaf shape and flowering times. However, it is possible that not all the complex variations observed in this species can be reduced to phenotypic variation superim- posed on one single species complex. Shull (1909) collected seeds of C. bursa-pastoris from different sites in North America, where the species is introduced but now widely naturalized, and grew these under standard conditions for several generations, by self-pollination. He found that the majority of his plants could be fitted into four basic groups based on the characters of the rosette leaf shape. He referred to these four groupings (“biotypes” sensu Shull) as Capsella bursa-pastoris and used these names for them: rhomboidea, simplex, heteris and tenuis. Steinmayer et al. (1985) examined 29 populations of C. bursa-pastoris from the Alps to northern Scandinavia, from Iceland and also a population from the Hindu Kush Mountains in Afghanistan, while Neuffer (1989) worked with populations of C. bursa-pastoris collected from southern to northern Europe (including three from Britain), two populations from Egypt and one from Israel. Their populations were grown under standard conditions either in glasshouses or in field trials for varying periods of time. Analyses of leaf shapes and capsule size allowed the authors to classify most of their plants into one of the four basic groups proposed by Shull (1909). This paper seeks to expand on the work of Neuffer (1989) by examining populations of C. bursa-pastoris from a variety of habitats and geographical areas in Britain, to determine how well Shull’s four basic groups are generally recognizable in the field, and from herbarium specimens; to germinate and grow seed from the different populations under standard conditions, and to observe whether, when environmental variation is removed, the plants can be classified according to Shull’s groups; and, finally, to determine whether or not this classification is maintained in their progeny. If Shull’s groupings are substantiated then a step will have been made towards simplifying the taxonomic classification of C. bursa-pastoris from 200 microspecies, or 25 segregates, or no attempt at all to sub-divide this variable species, to producing four useful groups, which can be recognized as having a distinct genetic basis underlying the environmentally modified phenotype. Although Shull’s work was concerned only with examining leaf morphology, other workers, mentioned above, have used capsule size and shape as identifying characters and these have also been examined in the present paper; chromosome numbers have also been assessed. Chromosome numbers are usually tetraploid with 2n = 32 (Davis 1965; Léve & Love 1956; Svensson 1983; Clapham et al. 1987). Chater (1993) records both 2n = 32 and also 2n = 16; Svensson (1983) also records diploid specimens with 2n = 16 from Greece. MATERIALS AND METHODS Seed samples from locally available populations were collected from 20 different habitats in Bradford and district, and from 14 other locations throughout Britain, between April and July 1993. CAPSELLA BURSA-PASTORIS 245 These were germinated in potting compost in a glasshouse and grown on until they set seed. Cross pollination was assumed to be prevented by keeping each of the populations in a different place in the glasshouse, with sliding separating doors to aid isolation, and by the fact that the species is primarily adapted for self-pollination. Subsequent seed collection and growing was continued through to the F, generation. In each generation the rosette leaf shape was assessed for 15 plants, randomly selected from each population, according to a four-fold categorisation on the basis of Shull (1909) as follows: Capsella bursa-pastoris group A (“‘simplex” sensu Shull (1909)) Leaves with mostly simple, rounded or triangular, acutish lobes. Capsella bursa-pastoris group B (“rhomboidea” sensu Shull (1909)) Leaves divided to the midrib; possessing a more or less rhombic terminal lobe, set off by deep sinuses from the nearest lateral lobes. Capsella bursa-pastoris group C (“‘heteris” sensu Shull (1909)) Leaves divided to the midrib; the terminal lobe usually separated from the nearest lateral lobes by deep, clean-cut incisions. Capsella bursa-pastoris group D (“‘tenuis” sensu Shull (1909)) Sinuses relatively shallow, rarely extending to the midrib; the terminal lobe is not separated from lateral lobes by deep incisions; lateral lobes are generally slender, elongated and acute. The original parent plants had also been evaluated on this categorisation. While most of these could be fitted to one of the groups a number were of intermediate status and could not be Classified at the time of collecting. For each of the F, plants, the lengths and widths of 30 seed capsules were measured to provide an indication of shape. The number of days elapsed between germination and the production of the first flower by each plant was recorded for each group. The somatic chromosome number was determined for 15 plants of the F, generation of each group from root-tip squashes. The root tips were pretreated for 2 hours in a solution of 0-1% colchicine and 2 mM 8-oxychinoline (1:1), fixed in Carnoy’s fluid and stained in aceto-orcein. In addition to the fresh material collected during this study, herbarium material from The Natural History Museum, London (BM) was consulted to obtain further information on the relative frequencies and geographical distribution of the groups in Britain, wherever the groups could be recognized from herbarium specimens. Two hundred specimens were examined, of which 189 could be allocated to one of Shull’s four basic groups. RESULTS Although the field populations showed phenotypic variation, the F, generation, produced by self-pollination, gave rise to F, and F, generations which exhibited no phenotypic variability from the F, generation. Observations and measurements from the F, generation are presented in Figs | & 2 and Table 1, divided into the leaf morphology groups shown consistently from the F, generation. The chromosome studies reported in this paper indicate that there are diploid as well as tetraploid populations of C. bursa-pastoris in Britain. Groups A and B (simplex and rhomboidea) are tetraploid while groups C and D (heteris and tenuis) are diploid. The results indicate that the rate of growth was greater for the tetraploid groups, which produced larger capsules (Fig. 2). Capsule TABLE 1. MEASUREMENTS OF VARIOUS PARAMETERS (MEAN + STANDARD ERRORS) FOR THE DIFFERENT GROUPS (A-D) (SENSU SHULL (1909)) OF CAPSELLA BURSA- PASTORIS GROWN UNDER STANDARD GREENHOUSE CONDITIONS Group A (simplex) B (rhomboidea) C (heteris) D (tenuis) Chromosome number 2n = 4x = 32 2n = 4x = 32 2n= 16 2n = 16 Capsule length (mm) 7-81 40-15 7:58 £ 0-15 6:56+0-11 6:29 + 0-11 Capsule width (mm) 7:34+0-11 6-47+0-11 4-85 + 0-07 4-25 + 0-06 Days to first flowering 32 + 4-26 40 +5-12 65 + 6:24 70 + 7:54 30 capsules from 15 plants for each group were measured. 246 A. AKSOY, W. H. G. HALE AND J. M. DIXON Ni rrr. mm FicureE 1. Comparison of leaf morphologies for F, Ficure 2. Comparison of capsule size and shape for F, Capsella bursa-pastoris groups A—-D, grown under Capsella bursa-pastoris groups A—D, grown under standard conditions: A - simplex; B - rhomboidea; standard conditions: A - simplex; B - rhomboidea; C - heteris; D - tenuis (sensu Shull (1909)). C - heteris; D - tenuis (sensu Shull (1909)). lengths for groups A (simplex) and B (rhomboidea), the tetraploids, were significantly different based on t-tests (at p < 0-001) from those of group C (heteris) and D (tenuis), the diploids. Capsule lengths were not significantly different between A and B nor between C and D. However, capsule widths were significantly different (at p < 0-001) between all four groups. Days to flowering was also examined and, as for capsule length, the differences between A and B were not significant, nor those between C and D, but the differences between the tetraploids and diploids were significant, based on t-tests, at p < 0-001 with the tetraploids flowering earlier (Table 1). Observations, however, suggest that the diploid groups tended to flower over a more extended period and to survive longer than the tetraploids. From observations of 289 individual plants (both herbarium and fresh) from Britain, the relative frequencies of the four groups were found to be: A (simplex) 23%; B (rhomboidea) 39%; C (heteris) 33% and D (tenuis) 5%. Based on this relatively restricted sample, the distribution of the four groups appears to vary geographically (Fig. 3); from the material (fresh and herbarium) available, groups B and C were found throughout Britain but group A was not recorded from northern Scotland and group D (based on a small percentage of records overall) was only found in material from England. DISCUSSION AND CONCLUSIONS The classification of C. bursa-pastoris in the field has been hampered by the polymorphic variation in leaf shape and size resulting from gene-environment interactions, particularly as the species is of such widespread geographical and altitudinal distribution and is found in a wide variety of habitats. Neuffer (1989), working with a large number of populations of C. bursa-pastoris, showed that the genotypes defining leaf type are easily modified by environmental parameters and that the degree of phenotypic plasticity varies from provenance to provenance. This variation has led to widely differing attempts to classify C. bursa-pastoris in terms of microspecies based on either leaf shapes and/or capsule shape and size. If systematic botany relies solely on descriptions of plants in the field then the problems of classifying polymorphic species such as C. bursa-pastoris will remain. However, if collateral cultivation methods are adopted in lieu of comparative field morphology, as has been carried out in this study, then a useful step forward may be made. Because genotype-environment interactions are very pronounced then the use of herbarium specimens in classifying C. bursa-pastoris, although also of considerable value, may have limitations owing to the presence of intermediate leaf shapes. It is not always possible to classify these into the basic groups, although in the current work only 5% of the herbarium specimens CAPSELLA BURSA-PASTORIS 247 5 — \ -- a 2 44° Rela poise iss Ce a Midi i det kL (BEE — Se Zeuae Figure 3. Di i a of Capsella bur 5 gro re s A- a Neve based on collected samples and herbarium mat val A simplex; B - rho rab ae as heteris; D - is (sensu Shull (1909)). 248 A. AKSOY, W. H. G. HALE AND J. M. DIXON examined could not be so classified. It nevertheless remains a possibility that some misclassifica- tion of specimens might have occurred because of phenotypic variation mimicking the characteris- tics of a different genetically-determined grouping. The glasshouse experiment carried out here shows that the original four-fold grouping of Shull, extended and defined by Steinmayer et al. (1985) and Neuffer (1989), also appears to hold good for British C. bursa-pastoris and although in the field a number of intermediates were collected, which could not be immediately classified into one of the four basic groups, on cultivation through to the F, generation all progeny derived from the intermediates could be placed into one of the four groups. The leaf morphology of these groups is easily recognisable and because these have a recognised genotype (Steinmayer et al. 1985) this seems a sensible sub-division on which to classify C. bursa-pastoris. Shull (1909) based his groupings only on leaf shape but in this paper capsule dimensions have also been examined and this additional work provides a further means of determining to which of the four groups individuals of C. bursa-pastoris belong. The differences in geographical distribution in Britain of the groups shown in Fig. 3 may possibly be a reflection of the restricted sample for which material has been examined and more sampling needs to be done to clarify this. However, differences in distribution of the groups, and in the relative abundance of the different groups, have also been suggested by other work elsewhere. Preliminary observations on plants in Turkey (not presented here) indicated that the same groups were recognisable there, although in that smaller sample only groups A, B and C were recorded. Group A was found only along the Mediterranean coast, while groups B and C were found both along the coast and also at inland sites. Neuffer (1989), in her more extensive studies throughout Europe, recorded 57% of her plants as belonging to group C (heteris), 41% as group B (rhomboidea) and 2% as group A (simplex). Group D (tenuis) again was not recorded. Confirma- tion of such differences would obviously be important for assessing the ecological significance of differential adaptations of these groups. The tetraploids (groups A and B) flowered earlier than the diploids (Table 1) under glasshouse conditions, as did those in the field trials of Neuffer (1989), and also those recorded during field observations by Svensson (1983) for C. bursa-pastoris in south-eastern Europe. This suggests that cross-pollination between the diploid and tetraploid groups is unlikely. Capsella bursa-pastoris flowers are adapted to both self- and cross-pollination although the former seems to be by far the more common method. This method is favoured by the fact that the stigma is receptive to pollen on both its under and upper surfaces and the anthers dehisce while they are held in contact with the under surface of the stigma by the partially-closed petals (Shull 1909). After the anthers begin to dehisce the petals open fully and pollen is available to be transported to other plants. Capsella bursa-pastoris is visited by a variety of small flies and bees (Shull 1909). Stebbins (1951) commented that self-pollinating annual species, in addition to showing genetic similarity, have a relatively high degree of phenotypic plasticity compared with perennials, so that the individual genotype can be modified considerably depending on whether the environment is favourable or not. It is the variable expression of this plasticity which has led to the confusion in the classification of C. bursa-pastoris. However, C. bursa-pastoris has not lost its capacity for forming new gene combinations and Hurka et al. (1989) estimated outcrossing rates, based on allozymes, as between 3 and 12%. They found that cold and rainy weather seems to support self-pollination, whereas dry and sunny weather seems to favour cross-pollination. Thus this predominantly autogamous, yet flexible mating system, coupled with polyploidy and extreme polymorphism, may contribute to the colonising ability of C. bursa-pastoris, and its widespread distribution. Shull did not present his names for the four types of Capsella bursa-pastoris as formal taxonomy. Indeed he states “‘in adopting the names used in this paper I have been governed entirely by the demands of present utility and not by any thought that these names will be accepted by taxonomists as having proper standing in the nomenclatural system now recognized by them” (Shull 1909). However, he goes on to record that a large number of plants of “the fourth pedigreed generation and a few of the fifth under observation” retained “easily recognized differentiating marks, which, except in one form, show no transgression of the characteristic features of any other form studied. These forms are therefore distinct elementary species, or biotypes, each characterized by certain constant features and each with its own range of fluctuating variability.” (Shull 1909). Shull CAPSELLA BURSA-PASTORIS 249 comments that he was at first inclined to use binomial names, leaving the Linnaean species name - bursa-pastoris - as the valid name for the aggregation of the infraspecific forms, having the same habit and triangular or obcordate capsules. But the fact that a corresponding series of infraspecific forms may occur in different related species led him to use the trinomial system, and he used the same infraspecific names for the two closely related species Bursa (= Capsella) bursa-pastoris and Bursa (= Capsella) heegeri. The authors are of the opinion that Shull’s (1909) interpretation of this taxon is based on sound characters underlying the complex phenotypic variability, and that consideration might be given to the establishment of subspecific nomenclature to describe these four groups of C. bursa-pastoris. ACKNOWLEDGMENTS We thank Mr S. Davidson for drawing the figures; Mr Iqbal for data on chromosome counts; the staff of The Natural History Museum, London and Ege University Herbarium, Turkey for providing access to herbarium materials; Dr E. C. Nelson, Professor M. R. D. Seaward and unnamed referees who commented on a draft of this paper; and the University of Erciyes for providing financial support for Dr A. Aksoy. REFERENCES Axsoy, A. (1996). Autecology of Capsella bursa-pastoris (L.) Medik. Ph.D. thesis, University of Bradford. ALMQUIST, E. (1907). Studien iiber Capsella bursa-pastoris 1. Acta Horti Bergiani 4: 1-92. AcmauisT, E. (1921). Bursa pastoris Weber. Supplement to report of the Botanical Society and Exchange Club of the British Isles 6: 179-207. ALMQUISsT, E. (1923). Studien iiber Capsella bursa-pastoris I\. Acta Horti Bergiani 7: 41-95. Cuater, A. O. (1964). Capsella, in Tutin, T. G., HEywoop, H., Burces, N. A., Moore, D. M., VALENTINE, D. H., Waters, S. M. & Wess, D. A., eds. Flora Europaea 1: 316. Cambridge University Press, Cambridge. Cuater, A. O. (1993). Capsella, in Tutin, T. G., HEywoop, H., BurGes, N. A., Moore, D. M., VALENTINE, D. H., Waters, S. M. & Wess, D. A., eds. Flora Europaea, 2nd ed. 1: 381-382. Cambridge University Press, Cambridge. CLAPHAM, A. R., TuTIN, T. G. & Moore, D. M. (1987). Flora of the British Isles, 3rd ed. Cambridge University Press, Cambridge. Davis, P. H. (1965). Flora of Turkey and the East Aegean Islands, 3. Edinburgh University Press, Edinburgh. Hou, G. L., PANCHO, J. V., HERBERGER, J. P. & PLUCKNETT, D. L. (1979). A geographical atlas of world weeds. Wiley-Interscience, New York. Hopxirk, C. P. (1869). A note on the forms of the genus Capsella. Bulletin de la Société Royale de Botanique de Belgique 8: 457. Hurka, H. (1990). Differentiation and adaptation in the genus Capsella (Brassicaceae), in KAWANO, S., ed. Biological approaches and evolutionary trends in plants, pp. 19-32. Academic Press, London. HurKA, H., FREUNDNER, S., Brown, A. H. D & PLANTHOLD, U. (1989). Aspartate aminotransferase isozymes in the genus Capsella (Brassicaceae): subcellular location, gene duplication, and polymorphism. Biochemi- cal genetics 27: 72-90. JorDAN, A. (1864). Diagnoses d’espéces nouvelles ou méconnues pour servir de matériaux a une Flore Réformée de la France et des Contrées Voisines, Vol. 1, Pt 1. F. Savy, Paris. Love, A. & Love, D. (1956). Cytotaxonomical conspectus of the Icelandic flora. Acta Horti Gotoburgensis 20 (4): 65-290. Manl, M. S. (1978). Ecology and phytogeography of high altitude plants of the northwest Himalaya. Oxford and Ibh Publications, New Delhi. MEIKLE, R. D. (1977). Flora of Cyprus. Royal Botanic Gardens, Kew. Motr, F. T. (1885). The Leicestershire forms of Capsella bursa-pastoris. The midland naturalist 3: 217-220. Neurrer, B. (1989). Leaf morphology in Capsella (Cruciferae): dependency on environments and biological parameters. Beitrdge zur biologie der Pflanzen 64: 39-54. Ricu, T. C. G. (1991). Crucifers of Great Britain and Ireland. Botanical Society of the British Isles, London. Rouy, G. & Foucaup, J. (1893). Flore de France, Vol 2. Société des Sciences Naturelles de la Charente- Inférieure, Asniéres. SHULL, G. H. (1909). Bursa bursa-pastoris and Bursa heegeri biotypes and hybrids. Carnegie Institution Publications 112: 1-57. 250 A. AKSOY, W. H. G. HALE AND J. M. DIXON Stace, C. A. (1989). Plant taxonomy and biosystematics, 2nd ed. Edward Arnold, London. Stace, C. A. (1997). New Flora of the British Isles, 2nd ed. Cambridge University Press, Cambridge. STEBBINS, G. L. (1951). Variation and evolution in plants. Columbia University Press, New York. STEINMAYER, B., WOHRMANN, K. & Hurka, H. (1985). Phanotypenvariabilitat und Umwelt bei Capsella bursa-pastoris (Cruciferae). Flora 177: 323-334. Svensson, S. (1983). Chromosome numbers and morphology in the Capsella bursa-pastoris complex (Brassicaceae) in Greece. Willdenowia 13: 267-276. WILson, A. (1949). The altitudinal range of British plants, 2nd ed. T. Buncle, Arbroath. (Accepted June 1998) Watsonia 22: 251-260 (1999) 251 Conservation of Britain’s biodiversity: Filago lutescens Jordan (Asteraceae), Red-tipped cudweed T. C. G. RICH Department of Biodiversity and Systematic Biology, National Museum and Gallery of Wales, Cardiff CF1 3NP ABSTRACT This paper summarizes the conservation work being carried out on Filago lutescens L. (Asteraceae), Red-tipped cudweed, a rare, statutorily protected species in Britain. It is a winter or spring annual, germinating mainly in the autumn, and flowering from June to October. Its habitats are mainly arable fields, tracks and path sides, open sandy ground, sand pits and commons or heathland usually in Thero-Airetalia vegetation. It has been recorded in a total of at least 212 sites in 86 10-km squares in south-eastern England, but has been seen in only 14 sites in ten 10-km squares since 1990. It appears to have declined in arable field habitats owing to changes in agricultural practices, and in tracks and heathland owing to reduced disturbance. Much of the decline took place before the 1960s. Population counts for all extant sites between 1993 and 1996 show marked variation from year to year and marked differences between sites. The best conservation management is currently thought to be annual disturbance by digging or rotavation in early autumn. This species is still under severe threat in Britain; only two extant sites have statutory protection and the two largest populations are unprotected. Keyworbs: population size, ecology, distribution, habitat management, rare species. INTRODUCTION Filago lutescens Jordan (F. apiculata G. E. Sm. ex Bab.; Asteraceae), Red-tipped cudweed, is a rare species in Britain. It is one of five species of Filago native to Britain all of which occur in open skeletal habitats, and all of which are declining. F. pyramidata L. is very rare and statutorily protected; its conservation is described by Rich (1999a). F. gallica L., recently reassessed as a Native species (Rich 1994), has subsequently been reintroduced from native stock to mainland Britain from where it has been extinct since the 1950s (Rich 1995a). F. vulgaris Lam. and F. minima (Sm.) Pers. have also shown significant declines between 1930 to 1960 and 1987 to 1988 (Rich & Woodruff 1996). Although F. lutescens has been known to be declining in Britain for over 20 years (e.g. Perring & Farrell 1983), virtually nothing was known about its ecology or the reasons for the decline. In 1994, the wild-plant conservation charity Plantlife became concerned that Filago lutescens, F. gallica and F. pyramidata were amongst the most threatened plants in Britain. The species were therefore included in their ‘Back from the brink’ project, which aims to conserve critically endangered plant species through research and management work. About 20 rare plant species have been included in this project between 1992 and 1996, which represents a significant contribution to the conservation of biodiversity in Britain by the voluntary sector. The aim of this paper is to summarize and update the conservation work carried out on F. lutescens up to 1996; full details can be found in Rich (1995b, 1996) and Rich & Davis (1996). Further details about the ‘Back from the brink’ project can be obtained from Plantlife. DISTRIBUTION DISTRIBUTION IN BRITAIN As Filago species have often been confused in Britain, a review of the historical records was first carried out. F. lutescens is relatively easily distinguished from the other species by the red tips to 252 T. C. G. RICH Ficure 1. Distribution of Filago lutescens in the British Isles. @ 1990 onwards. @ 1950-1989. O Pre-1950 and undated records. the young phyllaries — in older flowers these fade to straw-coloured, resulting in confusion with F. vulgaris and F. pyramidata. There are two subspecies of F. lutescens (Holub 1976): subsp. lutescens is widespread and occurs in Britain and Europe, and subsp. atlantica Wagenitz occurs in Portugal and the Azores. F. lutescens was first reported in Britain by Smith (1846) from sandy borders of fields, hedgebanks and roadsides in Yorkshire. Watson (1848) clarified the distinction between it and the other species, and it was subsequently quite widely recorded (not always correctly) in southern England. It has slowly declined, and is now very rare and unknown to most British botanists. Records have been compiled from the literature, herbaria (BM, BRISTM, BTN, CGE, E, GL, GLAM, IPS, K, LIV, LTR, MNE, NVW, NWH, OXF, RAMM, RNG, SLBI, TTN and US; abbreviations following Kent & Allen 1984), field survey, correspondence with botanists and from information held by English Nature and the Biological Records Centre, Monks Wood. Most of the records are supported by herbarium specimens determined by T. C. G. Rich or J. Holub. Doubtful records have been rejected. About 400 records have been traced, representing at least 212 localities in 86 10-km squares in 24 vice-counties. A distribution map distinguishing 1990 to 1996, 1950 to 1989 and older records is shown in Fig. 1. The species has been fairly widely recorded in south-eastern England. All the records are accepted as native as there are no records of any introduced localities. Note that a number of the records in Perring & Walters (1962, 1990) have been revised here and many more added. CONSERVATION OF FILAGO LUTESCENS 253 FicureE 2. Distribution of Filago lutescens (redrawn from Meusel & Jager 1992). The distribution of subsp. atlantica is enclosed by a dotted line. WORLD DISTRIBUTION The world distribution is shown in Fig. 2; it is widely recorded in western Europe, with scattered records in the former Yugoslavia (Holub 1976; Meusel & Jager 1992). It is still reasonably widespread in Spain, Portugal and France but becomes rarer eastwards in Europe. In Holland, it has been long-neglected and is probably extinct (Adema 1976), and there is no known reason for its disappearance (Mennema, Quene-Boterenbrood & Plate 1980). In Belgium and Luxembourg, it was recorded 18 times before 1930, and only once since (van Rompaey & Delvosalle 1972). In Germany, it is scattered and rare, mainly in the south and also in an area around the Baltic; there are relatively more pre-1945 records suggesting that it has declined (Haeupler & Schonfelder 1989). In Switzerland, there are 28 historical records mainly in the north-west of the country (Welten & Sutter 1982). In the Czech Republic, it is very rare and is statutorily protected. In the former U.S.S.R., it is treated as a variety of F. vulgaris, and is described as rare (Smol’ yaninova 1990). It is extinct in Sweden (Holub 1976). STATUS IN BRITAIN CHANGES IN THE NUMBER OF RECORDS WITH TIME A conservative approach has been taken to analysing the records to determine changes in the status with time, due to the inherent inconsistencies of the records and the recorders. Some records lack dates, some localities cannot be traced, and it is unclear if some records refer to the same or different sites. Whilst the general quality of most post-1950 records allows individual sites to be identified, this becomes increasingly difficult with older records. A site with the same name is assumed to be the same locality unless the habitats are obviously different (e.g. arable field or gravel pit). The absence of a record from a locality does not necessarily mean that the plant did not 254 T. C. G. RICH TABLE 1. SUMMARY ANALYSIS OF RECORDS OF FILAGO LUTESCENS IN BRITAIN BY DECADE. THE OVERALL NUMBER OF RECORDS IS 25% UNDER-ESTIMATED DUE TO UNDATED RECORDS Decade Total no. of No. of 10-km Extrapolated number of extant localities records squares per decade 1840—49* 26 12 21 1850-59* 5 5 7 1860-69* 11 6 8 1870-79* 15 12 16 1880-89 26 15 20 1890-99 29 16 25 1900-09 14 8 23 1910-19 24 15 23 1920-29 8 i, 14 1930-39 17 10 19 1940-49 9 6 16 1950-59 23 8 19 1960-69 20 10 17 1970-79 12 10 12 1980-89 17 3) 2 1990-1996 40 10 14 Undated 59 18 a2 * = Probably significant under-estimates due to lack of recording. Table 1 summarizes the records by decade, 10-km square and locality. Note that because a quarter of the historical sites have no date information, the number of records is under-estimated. The total number of records per decade varies according to the botanical activity of recorders rather than changes in the frequency of the plant, as is well known for other plants (e.g. Thlaspi perfoliatum; Rich 1999b). The publication of the first record in 1846 stimulated many other records from the Botanical Society of London, which was very active at the time, but the records drop in the 1850s when the Society collapsed. Collecting by its successor, the Botanical Exchange Club, in the 1860s and 1870s again resulted in many records with a peak in the 1890s, by which time the main features of the distribution were known. The troughs in the 1920s and 1940s may be due to the Depression and Second World War respectively. The influence of the agricultural revolution in the 1950s and 1960s is seen with a subsequent decrease in the number of records, the rise recently being due to field work for the Nature Conservancy Council in the late 1980s and the survey work in the 1990s reported below. The number of 10-km squares recorded is a widely-used measure of relative frequency in Britain (e.g. Perring & Farrell 1983). The number of 10-km squares recorded per decade shows a similar pattern (Table 1); note that because not all sites persist there is a considerable turnover of 10-km squares each decade. A better measure of change than the number of 10-km squares is the change in the number of sites with time. As not all sites will have been continuously recorded, the number of sites present has been extrapolated by assuming the species was present at each site for all decades between the first and last record. The number of sites present for each decade was then totalled (Table 1). There is a decline in the number of sites to the current 14 by gradual loss of sites randomly across the whole range, and most of the decline appears to have taken place by the 1960s. The plant has been recorded since 1990 in 14 out of at least 212 localities (6-6%), and many of the remaining populations are so small that they could be threatened in poor years (see below). Further losses are to be expected unless the remaining populations are conserved. REASONS FOR THE DECLINE The reasons for the decline have been investigated by analysing changes in the habitats in which the plant has been recorded (Table 2). Whilst some records do not indicate exactly in which habitat the plants were growing, the main trends can be discerned. CONSERVATION OF FILAGO LUTESCENS 255 TABLE 2. HABITATS OF FILAGO LUTESCENS IN BRITAIN COMPILED FROM RECORDS WITH HABITAT DATA. REPEATED RECORDS FROM THE SAME SITE ARE NOT INCLUDED. 114 RECORDS (43% OF ALL HISTORIC RECORDS) HAVE NO HABITAT NOTED Habitat Number (%) of records Number (%) of records up to 1989 1990-1996 Fields or arable 65 (43%) 4 (29%) Roadsides, lanes, paths, tracks 24 (16%) 3 (21%) Gravel and sand pits 12 (8%) 2 (14%) Commons and heathland 12 (8%) 1 (7%) Sandy or gravelly ground 12 (8%) 1 (7%) Fallow or stubble fields 7 (5%) 1 (7%) Railways 6 (4%) 1 (7%) Gardens 3 (2%) 0 Woods (presumably on tracks) 2 (1%) 0 Chalk pit 2(1%) 0 Meadow 1 (0-:7%) 0 Clay pit 1 (0-7%) 0 Golf links 1 (0-7%) 0 Rubbish tip 1 (0-7%) 0 Market garden 1 (0-7%) 1 (7%) Total 150 (100%) 14 (100%) Historically, Filago lutescens was most commonly reported from arable fields, associated with a wide range of crops including rye, corn, barley, wheat, clover, sainfoin, roots and potatoes (Table 2). Often the fields are described as sandy, and sometimes as gravelly. Interestingly, it was much less commonly reported from fallow or stubble fields than its two rare relatives, F. pyramidata and F. gallica, perhaps owing to the plant’s growing and flowering earlier than the other species. The decline in records from arable fields is striking. It was also widely reported from paths, tracks and roadsides (Table 2). Before roads were paved and covered with tarmac, road and tracks were regularly disturbed by carts and animals which must have created suitable habitats. Most road verges today tend to support unsuitable tall, closed grassland communities. Some unpaved tracks and paths still maintain suitable short, open habitats. The next most suitable habitats are sand or gravel pits, sandy ground, and commons and heathland. These reflect the suitability of the soils and the disturbance created by grazing or small-scale mineral extraction. The railway habitats include a range of sites from cuttings to tracks and station yards. The plant is only rarely recorded from other habitats, and has only occasionally been recorded from closed communities where it does not persist. The main reason for decline is undoubtedly the loss of populations in arable fields, probably due to the use of selective herbicides, fertilisers and changes in the timing of agricultural operations (Wilson 1992). This is a similar story to many other arable weeds in Britain such as Scandix pecten-veneris and Agrostemma githago. Loss of other habitats such as roadsides, and the decline in grazing and disturbance on heathlands, possibly coupled with the decline of rabbits through myxomatosis, have also taken their toll. POPULATION SIZES 1993 TO 1996 The population sizes of all post-1990 sites were monitored between 1993 and 1996 (Table 3; the population at Suffolk 5 was last seen in 1991). In addition, a considerable amount of time has been spent by many volunteers and the author searching old sites across the country (Rich 1995b). During this search one old site was rediscovered and two new sites found, and it is probable that there are still a few more undiscovered sites. The number of plants differs markedly between sites, with four sites holding on average over 90% of the plants (Table 3). The number of plants also varies markedly between years at some Sites, 256 T. C. G. RICH TABLE 3. ESTIMATED POPULATION SIZES OF FILAGO LUTESCENS AT ALL KNOWN SITES IN BRITAIN 1993-1996. SITES ARE LOCATED ONLY TO COUNTY Population size Site 1993 1994 1995 1996 Cambridgeshire 1 ‘frequent’ ‘frequent’ Unsurveyed Unsurveyed Essex 1] PX3) 59 20 7 Hampshire 1 114 135 26 1] Suffolk 1 10 0 0 0 Suffolk 2 30 3 0 0 Suffolk 3 1 seedling 0 0 0 Suffolk 4 - 5) 1000 5700 Suffolk 5 0 0 0 0 Surrey | 17 35 10 110 Surrey 2 2000 2000 2000 2000+ Surrey 3 2000 1000 5000 61 Surrey 4 1000 23142 No access No access Surrey 5 - : 300 100000 Sussex 1 200 500 40 20 Total 5397+ 26879+ 8400+ 107909+ Number of sites a 10 10 (+27) 8 (+27?) while staying relatively stable at others. Although some increases were responses to conservation management others were not, and there is no obvious link between population size and general management or climate; the only consistent pattern appeared to be an increase on arable fields in the second year of fallow. Were it not for exceptional increases in populations at two sites, the total British population in 1996 would be a matter of significant concern. ECOLOGY LIFE CYCLE Filago lutescens is an annual, as are all its close relatives. Observations in the field and in cultivation show that seeds germinate throughout the autumn, winter and spring (Rich 1995b). Moss carpets provided a good micro-site for germination in Essex, but plants were later more prone to drought (C. Gibson, pers. comm. 1994). Miiller (1995) studied germination at different temperatures, and found 100% germination at 15, 18, 20 and 22°C, without and with chilling (8 days in a freezer) treatments. Significant germination was found in all treatments within 6 days. Germination was a little slower at cooler temperatures, while treatment with Gibberellic acid resulted in enhanced rates of germination. No seeds in the dark at 15°C had germinated after 7 days, but once given light they germinated rapidly. Subsequently, seedlings not given the chilling treatment appeared to survive significantly better than seedlings given the chilling treatment, why is not known but it may reflect some aspect of the species’ ecology as it is noticeably western in distribution in Europe (cf. Fig. 2). Seeds thus seem to have little innate dormancy, and there is rapid germination of all viable seed when they grow in suitable conditions. There is unlikely to be a large, long-term seedbank. However, after clearance of Ulex europaeus scrub at the Hampshire site plants re-appeared after an absence of six years (C. Hall, pers. comm. 1995), and its appearance in arable fields in Surrey (J. E. Smith, pers. comm. 1996) also suggests that a small seed bank may be present in some sites. The plants over-winter as small rosettes, the stems elongating from about May onwards. The main flowering period is July to October. Plants in cultivation in Sussex flowered in the last week of June with F. vulgaris, a week after F. gallica and a week before F. pyramidata. On a still sunny July day, the heads of cultivated plants smelled sweet but no insects were observed to visit them. Plants are very variable in size in the field, perhaps related to time of germination and soil CONSERVATION OF FILAGO LUTESCENS 257 conditions. Plants observed severely droughted in July 1994 in Sussex were noted to have ‘greened up’ again in August 1994 after rains (C. Murray, pers. comm. 1994). Plants which had flowered in autumn 1993 in Essex were still present in December 1993 and over-wintered but died the following spring and did not set more seed (C. Gibson, pers. comm. 1994). VEGETATION AND SOILS Historically in Britain, F. Jutescens would have primarily been a species of weedy fields, but due to its decline in such habitats, it now appears to be a member of the annual communities of sandy, open places. The sites are usually disturbed and variable in composition. It is usually associated with species like Aira praecox, Myosotis discolor, Filago minima, F. vulgaris and strikingly, in most sites, Scleranthus annuus. In at least two British sites it has also been associated with both F. pyramidata and F. gallica. The vegetation is typically the Thero-Airetalia as it is in central Europe (Ellenberg 1988). This is a short-lived hairgrass community, one of a number of communities of heaths and grassland determined by human and animal activity. Other species characteristic of this vegetation type in Europe are Aira caryophyllea, Filago arvensis, Hypochaeris glabra, Moenchia erecta, Nardurus lachenalii, Ornithopus perpusillus (a weak associate), Scleranthus polycarpos, Teesdalia nudicaulis, Trifolium striatum, Tuberaria guttata, Vulpia bromoides and V. myuros. The species does not tolerate shade. Most of the sites have fine, sandy, well-drained soils. pH measurements ranged from pH (5-1-) 5-7 — 6-7 (-8-0), neutral to basic in terms of plant growth (Rich 1995b). Interestingly, in Hertfordshire, Webb & Coleman (1849) described it as occurring on “‘light but moister soil than its congeners’. All of the current sites are freely drained, and now only F. pyramidata can be found on moister clays. CLIMATE The distribution in south-eastern England suggests a requirement for warm summers and low rainfall. Most of the sites fall in the area with a mean daily July air temperature of above 16°C, and an annual rainfall of less than 800 mm (sites in Sussex and Hampshire may have up to 1000 mm a year). In Europe it is mainly distributed to the west of the 16°C July isotherm north to Denmark, with a sharper cut-off in distribution in central Europe and only minor extensions to the east (Meusel & Jager 1992). In wet years it is likely that plants will dampen off as happens with other Filago species. Seedlings over-watered in a greenhouse in spring 1995 damped off (P. Angold, pers. comm. 1995). Plants were observed to have survived snow for a period of at least one week during the winter of 1995/1996. Further observations are required. HERBIVORY Plants do not seem to be systematically grazed by rabbits, although the young inflorescences may be nibbled off, and they also tolerate minor damage from horse and cattle grazing. It is possible that the plants are distasteful and are avoided by large herbivores. Slug damage appears uncommon. CONSERVATION MANAGEMENT SITE MANAGEMENT It is essential that sites are appropriately managed each year to maintain open conditions as the plant appears to have only a limited seed bank. Traditionally, disturbance occurred during cultivation of the fields, but sites may now need rotavating on an annual basis. The best conservation management is currently thought to be annual disturbance by digging or rotavation in early autumn, but further research is required. Some of the current sites are unmanaged and the fact that the plants survive in some of them is amazing. Digging or rotavation of sites has been successful at the sites Sussex 1, Surrey 1 and Essex 1 (see Table 3), but attempts to resurrect ailing or recently extinct populations using this technique at Suffolk 1, 3 and 5 have failed. Some preliminary data for Surrey 3 suggest that in arable field situations harrowing is a better option than ploughing, presumably because it creates the open ground necessary without burying seeds too deeply in the soil. Clearance of Senecio jacobaea has been carried out at Sussex 1 and Suffolk 4 to eliminate the need for weed control by herbicide. The Surrey | site was nearly destroyed by unauthorized spraying of the protected roadside verge. 258 T. C. G. RICH An experiment designed to determine the best time for management work was carried out in the Sussex site in 1994-1995 by C. Murray, S. Dipper and T. C. G. Rich (Rich 1996). Replicated 1 m2 Squares were dug over by hand at monthly intervals from September 1994 to April 1995 and the number of plants counted in each plot in July 1995. Plants of F. lutescens appeared in plots dug over in all months. An analysis of variance showed that there were no significant differences (p = 0-07) between plots dug over in different months, indicating that the timing of management work is not critical. However, to ensure maximum population potential it is suggested that management work should be carried out in August or September depending on the season and occurrence of existing healthy populations. | Calculations have shown that a minimum sample size of 172 plants is required to preserve all, or very nearly all, polymorphic genes with frequency over 0-05 in a population (Lawrence, Marshall & Davies 1995a, b). It is thus proposed that conservation management should aim to achieve at least 172 F. lutescens plants at each site each year. On this basis, five sites have populations with long-term averages consistently above this minimum size but the others are distinctly threatened (Table 3). This species is still under severe threat in Britain. It is desirable that all populations of over 1000 plants, and at least one population in each county, should be given statutory protection to maintain local biodiversity. STATUTORY PROTECTION Filago lutescens is statutorily protected under Schedule 8 of the Wildlife and Countryside Act 1981 (as amended), which should prevent plants at all sites being picked or uprooted. The current protection of and threats to sites are summarized in Table 4. The Suffolk 2 and Surrey 3 sites are protected as statutory Sites of Special Scientific Interest (S.S.S.I.) specifically for F. lutescens, though it is currently extinct at one of these. Another S.S.S.I. designated for its bird populations has had two F. lutescens populations in the recent past (Suffolk 3 and 5). Surrey 1 and Sussex 1 have non-statutory nature conservation designations. The other seven sites, including the two largest, have no protection and are currently have no conservation management which is a particular cause for concern. MONITORING AND RESEARCH It is essential that populations are monitored each year, not only to establish new threats, but also to determine the results of the conservation work and assess natural variation due to weather. Between 1993 and 1996 monitoring was carried out cost-effectively by simple counts of plants, photographs and observations on management. Further research work into the factors controlling population size is urgently required. TABLE 4. SITE PROTECTION AND THREATS TO FILAGO LUTESCENS SITES IN 1996 Site Site protection and threats Cambridgeshire 1 No protection. Privately owned; owner aware of plant. Essex 1 No protection; threatened by fly-tipping and encroachment of vegetation. Management working. Hampshire | Local designation only; managed appropriately by local Wildlife Trust. Suffolk 1 No protection; appropriate management not working. Suffolk 2 S.S.S.1.; inappropriately managed. Suffolk 3 S.S.S.I.; appropriate management not working. Suffolk 4 No protection; co-operative owner. Suffolk 5 S.S.S.I.; appropriate management not working. Surrey 1 Protected road verge subject to regular catastrophes but otherwise sympathetically managed. Surrey 2 No protection; co-operative owner. Surrey 3 S.S.S.I.; appropriate management not working. Surrey 4 No protection and no access to assess threats. Surrey 5 No protection; horse grazed with potential threat of ‘improvement’ of grass. Sussex | Local designation only; co-operative owners S.S.S.I. = Site of Special Scientific Interest. CONSERVATION OF FILAGO LUTESCENS 259 EX-SITU CONSERVATION Seed from five extant sites and one extinct site is held at the Royal Botanic Gardens Seed Bank at Wakehurst Place (J. Terry, pers. comm. 1997). Collections from the other sites are urgently required to ensure conservation of the genetic diversity and allow for reintroduction programmes if needed. ACKNOWLEDGMENTS Many Plantlife members and other botanists have helped with this work, and I would like to thank Frances Abraham, Ken Adams, John Alder, Penny Angold, Malcolm Ausden, Margret Baecker, Nick Baxter, Gillian Beckett, Sarah Bird, Betty Bish, Chris Boon, Humphry Bowen, Paul Bowman, Lady Anne Brewis, Mary Briggs, Alec Bull, Rodney Burton, Clive Chatters, Barbara and Richard Chapman, Eva Crackles, Sue Dipper, Chris Durdin, John Edmondson, John Edwards, Lynne Farrell, Paul Fisher, Chris Gibson, Jenny Gill, Chris Hall, Nick Hinson, Alison Hoare, Enid Hyde, Trevor James, Stephen Jury, John Killick, John Knight, Paul Knipe, Peter Lawson, Phil Lusby, Roy Maycock, Andy McVeigh, Sigid Miiller, Clive Murray, Kath Owen, David Payne, Diane Rathel, Francis Rose, Mike Rowson, Martin Sanford, Ian Sanders, Pete Selby, Mike Shaw, Bill Shepard, Roy and Noreen Sherlock, Joyce Smith, Bill Sutherland, Janet Terry, Mike Thurner, Rebecca Warren, Keith Watson, Geoff Welch, Terry Wells, Joan Westcott, Martin Wigginton, Rupert Wilson, Dere Wise, Felicity Woodhead, Geoffrey Wilmore and Peter Youngs for their help. Peter Marren originally suggested the project. I would also like to thank the keepers of the herbaria named above for access to specimens. Roy Perry commented on the manuscript. The map was plotted using DMAP. I would also like to thank English Nature for access to their rare plant files. The work was funded by the Plantlife ‘Back from the brink’ project. REFERENCES ApEMA, F. (1976). De Filago vulgaris-groep in Nederland. Gorteria 8: 41-47. ELLENBERG, H. (1988). Vegetation ecology of Central Europe, 4th ed. Cambridge University Press, Cambridge. Everett, S. (1987). Rare plant survey of South Region. Confidential report to N.C.C., Peterborough. HaEupLeR, H. & SCHONFELDER, P. (1989). Atlas der Farn- und Bluten-pflanzen der Bundesrepublik Deutschland. Ulmer, Stuttgart. Ho.us, J. (1976). Filago L., in T. G. Tutin et al., eds. Flora Europaea 4. Cambridge University Press, Cambridge. JERMYN, S. T. (1974). Flora of Essex. Essex Naturalists’ Trust Ltd, Colchester. KENT, D. H. & ALLEN, D. E. (1984). British and Irish Herbaria, 2nd ed. B.S.B.I., London. Lawrence, M.J., MARSHALL, D.F. & Davies, P. (1995a). Genetics of genetic conservation. I. Sample size when collecting germplasm. Euphytica 84: 89-99. Lawrence, M.J., MarsHALL, D.F. & Davigs, P. (1995b). Genetics of genetic conservation. II. Sample size when collecting seed of cross-pollinating species and the information that can be obtained from evaluation of material held in gene banks. Euphytica 84: 101-107. MENNEMA, J., QUENE-BOTERENBROOD, A. J. & PLATE, C. L. (1980). Atlas of the Netherlands flora. 1. Extinct and very rare species. W. Junk, The Hague. MEusEL, H. & JAcgr, E. J. (1992). Vergleichende Chorologie der Zentraleuropaischen flora. Gustav Fischer Verlag, Jena. Stuttgart. MU Ler, S. (1995). Seed germination and dormancy breaking. B.Sc. thesis. University of Sussex, Brighton. PERRING, F. H. & FARRELL, L. (1983). British red data books, 1. Vascular plants. R.S.N.C., Lincoln. PERRING, F. H. & Watters, S. M. eds (1962). Atlas of the British flora. B.S.B.1., London. PERRING, F. H. & Watters, S. M. eds (1990). Atlas of the British flora. 3rd ed. B.S.B.I., London. Ricu, T. C. G. (1994). Narrow-leaved cudweed (Filago gallica) in Britain. Back from the brink project report no. 22. January 1994. Plantlife, London. Ricu, T. C. G. (1995a). Extinct is forever... unless you happen to be very lucky! Sanctuary 24: 35. _ Ricu, T. C. G. (1995b). Red-tipped cudweed (Filago lutescens) in Britain. Back from the brink project report no. 28. January 1995. Plantlife, London. Ric, T. C. G. (1996). Red-tipped cudweed (Filago lutescens) in Britain in 1995. Back from the brink project report no. 68. January 1996. Plantlife, London. Ricu, T. C. G. (1999a). Conservation of Britain’s biodiversity IV: Filago pyramidata, Broad-leaved cudweed. Edinburgh journal of botany (in press). 260 T. C. G. RICH Ricu, T. C. G. (1999b). Conservation of Britain’s biodiversity I: Thlaspi perfoliatum L. (Brassicaceae) Cotswold Pennycress. Biodiversity and conservation (in press). Ricu, T. C. G. & Davis, R. (1996). Red-tipped cudweed (Filago lutescens) in Britain in 1996. Back from the brink project report no. 84. November 1996. Plantlife, London. Ricu, T. C. G. & Wooprurr, E. R. (1996). Changes in the floras of England and Scotland between 1930-1960 and 1987-1988: The B.S.B.I. Monitoring Scheme. Biological conservation 75: 217-229. Smith, G. E. (1846). Botanical Society of London Report. Phytologist 2: 575-576. SMOL’YANINOVA, L. A. (1990). Flora of the USSR 25: 336-349. ed. B. K. Shishkin. (English Translation) Koeltz Scientific Books, Koenigstein. VAN RompaAEY, E. & DELVOSALLE, L. (1972). Atlas de la flore Belge et Luxembourgeoise. Jardin Botanique National de Belgique, Brussels. Watson, H. C. (1848). Some account of the alleged species included under the name of Filago germanica Linnaeus. Phytologist 3: 313-318. Wess, R. H. & CoLeman, W. H. (1849). Flora Hertfordiensis. William Pamplin, London. WELTEN, M. & Sutter, H. G. R. (1982). Atlas de distribution des Pteridophytes et des Phanerogames de La Suisse. Birkhauser, Basel. WItson, P. J. (1992). Britain’s arable weeds. British wildlife 3: 149-161. (Accepted June 1998) Watsonia 22: 261-267 (1999) 261 Juniper in the Lake District National Park. A review of condition and regeneration T. C. DEARNLEY 50 West Street, Tetbury, Gloucestershire, GL8 8DR and J. G. DUCKETT Queen Mary & Westfield College (University of London), Mile End Road, London, E1 4NS ABSTRACT Juniperus communis L. (Common Juniper) (Cupressaceae) is a native British species of evergreen dioecious conifer, threatened by extensive grazing, competing tree species and lack of sites to colonise. This study assesses the present status of juniper in the Lake District. Ten large stands recorded as in good condition in 1975 were compared to five smaller stands, and a reference stand protected from grazing for 70 years. Recorded values of the number of berries produced by large stands and seed viability of these berries were combined as a seed viability index. Analysis showed that the seed viability indices of large stands were significantly higher than the small stand values, but significantly lower than the reference stand. These results suggest low reproductive potential may be indicative of a senescing population, and that grazing pressure is limiting reproduction. The absence of regeneration is attributed to stands becoming substantially moribund at a similar time without replacement. Seedling propagation and planting in fenced areas is suggested as the best strategy for juniper conservation. KEYWORDS: reproductive potential, seed viability index, stand size, grazing pressure. INTRODUCTION Juniperus communis L. is a characteristic shrub of varied morphology. It is one of three native species of Coniferae in the British Isles (Humphries 1981), well established in the Lake District where it is mainly found on scree slopes and exposed regions. Throughout the country J. communis exhibits variation in population structure existing sometimes as large stands, sometimes as individual bushes, and has been observed to show differences in life history, genetic constitution and morphology, particularly between southern and northern Britain. The north/south divide is also apparent from the distribution map of the species (see Fig. 1). Two subspecies of J. communis are recognised, J. communis subsp. nana Syme which is usually prostrate, small and slow growing, and J. communis subsp. communis. the commoner, larger subspecies. However there is some doubt as to the distinction between subspecies due to the highly variable morphology (Ward pers. comm.). Evidence from Upper Teesdale Nature Reserve, North Pennines and Tynron Juniper Wood, Dumfries and Galloway shows that there has been active management of J. communis in this country for a considerable period (Piggott 1956). By continuously disturbing the ground over a large area, substantial quantities of J. communis have been maintained which would otherwise have been outshaded by successional species such as Betula pendula (Clifton, Ranner & Ward 1997). In the Lake District particularly, the wildly fluctuating mining and quarrying industries which were a feature of the area for several hundred years up until this century (Gilbert 1980), and the lack of any regulated grazing regime are thought to have encouraged colonisation by continuous disturbance of the ground (Milner 1992). The large quantity of J. communis in the Lake District is therefore thought to be due to two main influences: (1) active management and (2) conditions which are conducive to repeated colonisation. 262 DEARNLEY AND DUCKETT JUNIPERUS COMMUNISL. ’ Juniper e 1930 onwards © Before 1930 Records from other counties probably Introductions Sages ae TM mM OSS Le | Figure 1. National distribution of Juniperus communis The vast majority of research into British J. communis has been carried out in Southern England. As far as we are aware there are no published papers relating to J. communis in the Lake District, and while Miles & Kinnard (1979a; b) have researched J. communis in the Scottish Highlands, and Ward has made maps of distribution over the Lake District (Ward pers.comm.), there remains a gap in our knowledge of J. communis in this region. It is recognised however that J. communis is declining in the area and has been for most of this century, as it is over the majority of the country (Ward 1973). There has been speculation over the reasons for the present decline in J. communis but Ward (1977), Gilbert (1980) and Clifton, Ranner & Ward (1997) have all suggested that lack of suitable sites to colonise and intensive grazing of any regeneration are the principal causes. Ward (1982) has conducted surveys of the age of J. communis and results show that most stands have a very even age structure. This is taken as evidence of the colonising nature of J. communis and also suggests that without further colonisation, individual stands will tend to become universally moribund at a particular age. Evidence that J. communis declines in seedling viability with age of parent means that if stands are allowed to become moribund, the potential for natural regeneration will also decline (Raatikainen & Tanska 1993). In the Lake District particularly, the grazing and mining regimes used until the end of the last century, are thought to have been beneficial for recolonisation. At present any mining is done on a JUNIPER IN THE LAKE DISTRICT 263 more long term basis and is less disturbing to the local community. The widespread change of farming practice from arable to livestock following the agricultural depression of 1870—1940 seems to coincide with the start of the J. communis decline in the area (Milner 1992). Grazing would appear to play an important role in J. communis colonisation, as noted by Thomas (1960; 1963) when the widespread outbreak of myxomatosis in the 1950s was followed by an unprecedented period of regeneration. Rabbit grazing seems to bare the ground sufficiently to allow J. communis growth, if the grazing pressure from rabbits and other herbivores is then reduced. However Fitter & Jennings (1975) found that the removal of grazing altogether allows other species to overshadow J. communis seedlings within three years. Thus some grazing appears to be essential. Several other suggestions have been made as to the conditions necessary for regeneration, including fire to bare the ground, death of other plants, trampling by farm animals and disturbance by moles (Miles & Kinnard 1979a; 1979b), which all assist the colonising nature of J. communis. Few experiments have been conducted to test these theories, with the exception of sheep grazing experiments by Fitter & Jennings (1975). This paper presents the results of a 1995 survey of J. communis in the Lake District. The aims were (1) to assess the status and (2) to propose recommendations for future management. As this survey did not record variables over a period of time, substantial sites of juniper were compared with smaller sites, to discover whether there were differences in seed viability and berry abundance. Results of the present survey are compared with others carried out in 1975 (Ward pers. comm.) and in 1995 (Sear 1995), which had similar aims to the present study. METHOD Ten substantial stands > 1000 of J. communis were chosen which were in good condition in 1975. As these large stands were recognised as worth conserving in 1975, but had not been actively managed since this time, any change in their condition was likely to be apparent in September 1995 when the survey was conducted (see Table 1). The ten large stands were compared with five smaller stands of < 1000 bushes (small stands). The stands selected for this survey encompass the range of habitats occupied by J. communis throughout the Lake District (Table 1). An area of J. communis which had not been recorded in 1975 known as Juniper Scar was also examined. This had been documented for several years and more importantly, surrounded by a sheep proof fence for the last seventy years. This stand was therefore examined to assess how protection from grazing might influence the ecology of J. communis. Three 100 m’ quadrats per stand were examined. The quadrats were distributed over the stand at three points, representing the variations in physical condition and exposure of the site. Notes were taken on the general appearance of each stand, and appearance of individual quadrats. Perhaps the most useful variable which could have been recorded in this survey was the age of stands. However J. communis is a notoriously difficult species to age without cutting live samples of the stem, for two reasons. The first is that the stem diameter of J. communis is not closely related to age. The second is that J. communis stems are usually eccentric in shape and therefore their girth is difficult to record with any accuracy; this also precludes accurate core sampling (Fitter & Jennings 1957): Seed viability has obvious implications for the reproductive potential of a stand of J. communis, and is therefore an important indicator of present and future condition which may not be obvious to the eye. Ten mature, purple berries were collected in each quadrat using random number tables to select the nearest female bush to given co-ordinates. Viability was determined by sectioning berries. Seeds are viable “where the internal tissues fill the seed completely and are white/off white in colour’ (Ward 1989). Viability was then recorded as a mean of the three samples out of a possible thirty seeds. A seed production index was obtained using the abundance of berries on bushes at each site. This was calculated by multiplying the number of bushes with a limited number of berries by one, the bushes with abundant berries by ten, and the bushes with very abundant berries by one hundred. Together these values gave the mean seed production index for each site (from Sear 1995). The seed viability index is the mean proportion of a possible 30 seeds at each site which were viable, multiplied by the seed production index to show how many of the seeds produced were viable, as sites may have produced many berries but these may have contained very few viable seeds. This value demonstrates the condition of the stand and the likely regeneration potential, as far 264 DEARNLEY AND DUCKETT TABLE 1. SITES STUDIED WITH GRID REFERENCES Site Grid Reference Habitat Altitudem Geology Large sites Place Fell NY/396.180 Rough fell 210 Volcanic Mardale Banks NY/482.124 Grazing 290 Volcanic Carrock Fell ‘NY/329.327 Grazing 310 Volcanic Dovedale NY/380.116 Scree 610 Volcanic Blind Tarn Moss NY/314.070 Fell 270 Volcanic Blea Tarn NY/295.037 Grazing 350 Volcanic Thwaites Fell SD/177.904 Grazing 190 Volcanic Yew Barrow SD/354.871 Woodland 230 Volcanic Bradleyfield SD/489.921 Grazing 190 Carboniferous limestone Whitbarrow SD/442.893 Grazing 200 Carboniferous limestone Small sites Whitbarrow N. SD/443.889 Fell 140 Carboniferous limestone Lingmore Oak NY/300.057 Fell & grazing 190 Volcanic Broad Hollins SD/299.914 Bog 150 Valley bog Blea Tarn S. NY/298.836 Fell 320 Volcanic High Harsop NY/393.105 Fell & grazing 410 Volcanic Reference site Juniper Scar NY/476.012 Nature reserve 260 Volcanic (Staveley Head Fell) TABLE 2. SEED VIABILITY INDICES OF JUNIPERUS COMMUNIS IN THE THREE SITE CATEGORIES Site Seed Viability Seed Viability Index Large sites Place fell 14-00 29-73 Mardale Banks 15-00 83-84 Carrock Fell 16-67 200-25 Dovedale 12-33 9-86 Blind Tarn Moss 15-33 5-47 Blea Tarn 15-00 2-65 Thwaites Fell 14-33 6-05 Yew Barrow 9-33 0-52 Bradleyfield 10-00 18-67 Whitbarrow 11-67 3-90 Small sites Whitbarrow N. 5-33 1-24 Lingmore Oak 15-00 13-49 Broad Hollins 11-33 5-67 Blea Tarn S. 11-67 4-67 High Harsop 17-33 7-32 Reference site Juniper Scar 17-33 243-39 (Staveley Head Fell) JUNIPER IN THE LAKE DISTRICT 265 @Reference site CiLarge sites @WSmall sites Mean number of bushes per quadrat @ 4 2 0 None Lim ited Abundant Very abundance abundant Berry abundance on bushes FicureE 2. Berry abundance on bushes at each site category. None - No berries. Limited abundance - Few berries on some branches. Abundant - Berries on most branches, some in clumps. Very abundant - Bushes laden with berries, most in large clumps. 250 200 150 100 50 Mean seed viability index 0 Reference Large sites Sm ali sites site Site category FicureE 3. Mean seed viability of Juniperus communis L. at each site category. as seed production and viability are concerned. Values are recorded in Table 2. The seed viability index for each site category were then compared using a one-way ANOVA test. RESULTS The null hypothesis that there is no difference in seed viability index between the site categories was rejected at the 5% significance level. The average seed viability indices were: Small Sites = 6-478, Large Sites = 36-094 and Reference Site = 243-39. The calculated F value was greater than the tabulated F value as F.,,, = 8-6063 > F.,, = 4-75 for 2 degrees of freedom between groups and 13 error degrees of freedom within groups. As the large sites had significantly higher seed viability indices than the small sites, there is a > 95% probability that large sites produce more viable seeds than small sites. Large sites also produce significantly less viable seeds than the reference site (Juniper Scar). Berry abundance and seed viability indices at each category of site are illustrated in Figs 2 & 3 respectively. In addition to the 16 sites recorded in this survey, an example of attempts to encourage J. communis regeneration was visited at Greenside Mines. Although the ground was heavily 266 DEARNLEY AND DUCKETT polluted with lead, J. communis seedlings were growing well within protective tubes. However, these will have to remain protected for at least eight more years before they are able to resist grazing (Ward & Lakhani 1977), particularly from sheep which occupy the surrounding land. This regeneration experiment provided a useful example of how conservation measures could be implemented to propagate J. communis in the Lake District. DISCUSSION The seed viability indices found in this survey are consistent with data obtained by Ward (1989) at Teesdale National Nature Reserve, which showed that the maintenance of a high seed viability is important for the conservation of J. communis. This suggests that the sites recognised as good in 1975 producing adequate quantities of berries, are still in a more reproductively viable condition than the small sites in this survey. However the much higher viability index recorded at Juniper Scar may be an indication of the seed reproductive potential and berry production necessary to maintain regeneration, as recommendations for the maintenance of viability such as disturbance and fencing (Miles & Kinnard 1979a) are found at this site. In this case the ten large sites may be in worse condition than in 1975, but have experienced a viability decline in parallel with the small sites. Without long-term experimental evidence that seed viability and berry production correlates with regeneration, it is not possible to suggest a limit at which regeneration is endangered, but the evidence from these three categories of sites would seem to suggest a decline over the last 20 years © at least. CONSERVATION AND FUTURE RECOMMENDATIONS The situation at Greenside mines is characteristic of the problems of conservation schemes in the Lake District. Such is the need for plant cover on steep slopes of the area, that the National Parks have sometimes had to implement emergency action to stop intense erosion (Harding, pers. comm.), but long term schemes are often restricted by land ownership complications. Several authors have drawn attention to the need for long term surveys to be conducted into the causes of declining J. communis in Britain including Ward (1987) and Ranner (1994), who also used Ward’s 1975 survey as a baseline for his own study. With the exception of regeneration experiments at Teesdale National Nature Reserve (Findley, pers. comm.) which are at present unpublished, there have been few studies which have been conducted thoroughly enough and over a long enough period to suggest techniques which may be employed to encourage regeneration. The problem with long term experiments at present is that the decline of J. communis may be extremely severe before conclusions can be put into practice. Also, the range of morphologies and possibly genealogy of J. communis in the Lake District alone, may mean that J. communis varies in the conditions needed for regeneration between regions, and possibly between stands. Milner (1992) underlines that J. communis propagation should use cuttings from local sources as the species has a wide genetic diversity. Although historical aerial or terrestrial photographs were not used for our study, this source of information on past populations and colonisation could be invaluable for rapidly establishing changes in population density in other regions. As intense grazing seems to be one of the central causes of J. communis decline, controlled grazing regimes should be put into practice as soon as possible, perhaps using the collapsible fences recommended by Miles & Kinnard (1979a). These authors suggest other means of encouraging regeneration such as felling mature trees and sowing seed, but these are experimental techniques not yet investigated for the conservation of J. communis. Propagating seedlings and planting in fenced areas seems to be the best immediate approach to J. communis conservation. The National Parks Centre at Brockhole have found that propagating cuttings is an effective way to grow stock. Roughly 60% of Greenside Mines cuttings were successfully grown in greenhouse conditions (Tasker, pers. comm.), which also preserves genetic diversity. Long term studies of reproductive condition in parallel with planting are urgently required. Unless these are instigated, young or small populations of J. communis face great danger through population fluctuation in the near future. JUNIPER IN THE LAKE DISTRICT 267 ACKNOWLEDGMENTS Tom Dearnley gratefully acknowledges a grant from English Nature. The authors would like to thank Joanna Backshall, Stephen Cull, John Day, Ian Findley, Roy Harding, Sue Tasker and particularly Lena Ward for practical guidance and comments on the manuscript. Opinions expressed in this report are those of the authors alone. REFERENCES CuiFTon, S. J., RANNER, D. S. & Warp, L. K. (1997). The status of juniper Juniperus communis L. in north-east England. Biological conservation 79: 67-77. Fitter, A. H. & JENNINGS, R. D. (1975). The effects of sheep grazing on the growth and survival of seedling juniper (Juniperus communis L.). Journal of applied ecology 12: 637-642. GILBERT, O. L. (1980). Juniper in Upper Teesdale. Journal of ecology 68: 1013-1024. Humpnrries, C. J., Press, J. R. & Sutton, D. A. (1981). Trees of Britain and Europe. Reed, Hong Kong. Mites, J. & KINNARD, J. W. (1979a). Grazing with particular reference to birch, juniper and Scots pine in the Scottish highlands. Scottish forestry 33: 102-117. MILEs, J. & KINNARD, J. W. (1979b). The establishment and regeneration of birch, juniper and Scots pine in the Scottish highlands. Scottish forestry 33: 280-289. Miner, J. E. (1992). The tree book. Collins & Brown, Oxford. Morris, M. G., CLARKE, R. T., RisPiIn, W. E., Warp, L. K. & THomas, J. A. (1993). Invertebrate studies at Old Winchester Hill NNR. Contract report English Nature/Natural Environmental Research Council. Unpublished report held by English Nature/Institute of Terrestrial Ecology. Picotr, C. D. (1956). The vegetation of Upper Teesdale in the North Pennines. Journal of ecology 44: 545-586. RAATIKAINEN, M. & TansKA, T. (1993). Cone and seed yields of the juniper (Juniperus communis) in southern and central Finland. Acta botanica sennica 149: 27-39. Ranner, D. S. (1994). Juniperus communis. History, current status and future conservation, with particular reference to County Durham. H.N.D. Environmental Land Management. Newton Rigg College. Unpublished. RopweELL, J. S. ed. (1991). British plant communities 1. Woodland and scrub. Cambridge University Press, Cambridge. Sear, J. (1995). Juniper in the Lake District National Park. Condition and regeneration. B.Sc. dissertation, University of Lancaster. Tuomas, A. S. (1960). Changes in vegetation since the advent of myxomatosis. Journal of ecology 48: 287-306. Tuomas, A. S. (1963). Further changes in vegetation since the advent of myxomatosis. Journal of ecology 51: 151-186. Warp, L. K. (1973). The conservation of juniper: Present status of juniper in southern England. Journal of applied ecology 10: 165-188. Warp, L. K. (1977). The conservation of juniper: The associated fauna with special reference to southern England. Journal of applied ecology 14: 81-120. Warp, L. K. & LaKHANI, K. H. (1977). The conservation of juniper: the fauna of food-plant island sites in southern England. Journal of applied ecology 14: 121-135. Warp, L. K. (1982). The conservation of juniper: longevity and old age. Journal of applied ecology 19: 917-928. Warp, L. K. (1987). Management of grassland - scrub mosaics. Reserve management 33: 134-139. Warp, L. K. (1989). Seed viability in juniper (Juniperus communis). Contract report English Nature/Natural Environmental Research Council. Unpublished report held by English Nature/Institute of Terrestrial Ecology. (Accepted June 1998) Watsonia 22: 269-273 (1999) 269 Correlations between higher-taxon richness and species richness in the British Isles flora D. M. WILKINSON Biological and Earth Sciences, Liverpool John Moores University, Byrom Street, Liverpool, L3 3AF ABSTRACT It has been suggested that the biodiversity of areas may be compared using the number of higher taxa present rather than by compiling full species lists. This would reduce sampling problems especially in the tropics with their high species richness. Data from British Isles regional Floras are used to examine the relationship between species richness and higher taxa richness for flowering plants. Positive correlations were found between species richness and genus richness (r,= 0-97) and species richness with family richness (r= 0-89). Richness of species in a small number (15) of the most species-rich genera was highly correlated with total species richness (r, = 0-91) and could be used as an alternative to higher taxa in comparing sites. Such relationships, validated on the well studied British flora, could be of great use in less well studied areas such as tropical forests. The study also illustrates the importance of regional Floras as a data base for use in studies of biodiversity. KEyworbDs: regional Floras, biodiversity, higher taxa, species richness. “The ultimate task of the systematist is not merely to describe the diversity of the living world but also to contribute to its understanding.” Mayr (1997). INTRODUCTION One of the key problems of conservation biology is how to identify the most important areas for biodiversity so that they can receive protection (Williams & Gaston 1994; Wilson 1992). An ideal data set on which to base such a decision would include full species lists for all areas under discussion. However, even for biologically well known countries such as Britain, this has not been achieved for many groups. The problem is much greater in the tropics, the home of the greatest terrestrial biodiversity (Wilson 1992); here estimates of the total number of species are very large. For example, estimates of the number of insects based on extrapolation from beetle data (reviewed by May-(1988) and Gould 1996)) give values of up to 50 million species. Studies based on extrapolation from other insect orders give smaller, but still very large numbers (Hodkinson & Casson 1991). Whatever the true number, it is clear that full species lists will not be achieved, even for the apparently less diverse flowering plants, which have an estimated 250,000 species (Holdgate 1991). This raises the question of how areas can be compared when full species lists are not available. A number of approaches have been described, including: i. use of correlation between biodiversity and environmental variables (e.g. soil or water chemistry); ii. relationships between indicator groups and total biodiversity; iii. using higher taxa to compare sites (Williams & Gaston 1994); and iv. combining data from groups of widely differing organisms (Vane-Wright et al. 1994). This paper concentrates on the use of higher taxa, based on the idea that relationships between species richness and number of higher taxa can be used to compare sites. Such an approach has been widely used in palaeontology, where making comparisons at the species level is often difficult (Gaston & Spicer 1998). This approach, if successful, would be in many cases logistically more realistic and more cost-effective than compiling full species lists (Andersen 1995; Gaston & Williams 1993; Williams & Gaston 1994). This paper aims to investigate the relationship between species richness and higher taxon (family 270 D. M. WILKINSON TABLE 1. NUMBERS OF FAMILIES, GENERA AND SPECIES RECORDED IN 20 BRITISH AND IRISH REGIONAL FLORAS Number of: families genera species species in Area Reference large genera Suffolk Simpson, 1982 | 118 622 1643 221 Kent Philp, 1982 119 657 1643 216 Dorset Good, 1984 110 585 1473 225 Glamorgan Wade et al., 1994 126 570 1460 Dis Jersey Le Sueur, 1984 120 585 1448 201 Durham Graham, 1988 113 546 1421 220 Somerset Roe, 1981 110 Sie 1413 223 Angus Ingram & Noltie, 1981 105 491 1192 212 Leicestershire Primavesi & Evans, 1988 98 505 1171 189 East Yorkshire Crackles, 1990 104 476 L338 184 Shropshire Sinker et al., 1985 110 475 1089 186 Derbyshire Clapham, 1969 101 478 1089 186 South Lancashire Savidge et al., 1963 103 479 1063 185 Staffordshire Edees, 1972 98 436 1017 Wil Anglesey Roberts, 1982 102 450 1004 188 Cheshire Newton, 1971 100 427 913 P52 Radnorshire Woods, 1993 99 387 871 liv) Connemara & Burren Webb & Scannell, 1983 94 387 818 160 Outer Hebrides Pankhurst & Mullin, 1991 81 321 696 148 Shetland Scott & Palmer, 1987 79 300 627 93 and genus) richness, for the well studied British flora, to test the correlation between species and higher taxa richness. It also considers the importance of the more species-rich genera in such relationships. METHODS The long history of biological recording in Britain, often carried out by non-professional biologists (Allen 1976) makes it one of the best studied countries in the world. This provides data sets which can be used to test relationships between species richness and higher taxon richness. This study used data on flowering plants from a subset of the regional Floras produced for Britain and Ireland (Table 1), most of which cover a single Watsonian vice-county. This system dates from 1852 and divides Britain into 112 vice-counties (Ireland is covered by an additional 40). The idea was to create a set of areas of more equal size than the administrative counties, into which the country could be divided for biological recording. These vice-counties are still somewhat unequal in size with between three and six per 100 km? (Vincent 1990). While such sizes are large by the standard of British reserves they are more comparable in size with some existing and proposed tropical protected areas (Janzen 1994; Wilson 1992). In 20 such regional Floras the numbers of species, genera and families of flowering plants were counted. As Ratcliffe (1984) pointed out, humans have introduced a large number of non-native species to Britain and altered the distribution of many native plants. No attempt was made to distinguish native from non-native species, all species listed in a given Flora as growing in the area were counted. The coverage of apomictic microspecies of Rubus and Taraxacum is very variable between different Floras, and therefore in this study they were counted as two species R. fruticosus agg. and T. officinale agg. The apomictic microspecies of the genus Sorbus were also excluded from the calculations. Such microspecies are really clonal lineages and can be regarded as single HIGHER-TAXON RICHNESS AND SPECIES RICHNESS 271 TABLE 2. LARGE GENERA, DEFINED AS ONES WITH 20 OR MORE SPECIES IN BRITAIN AND IRELAND LISTED BY STACE (1991) Excluding Sorbus with its many apomictic microspecies Genus Number of species Carex 74 Cotoneaster 45 Salix 35 Trifolium 32 Ranunculus 31 Potamogeton 29 Juncus 28 Chenopodium 27 Veronica 26 Geranium 24 Rumex 23 Saxifraga 22 Euphrasia 21 Allium 20 Rosa 20 evolutionary individuals rather than good species (Janzen 1977); as such they should probably not be considered as important as other species in biodiversity surveys (see Gould (1996) for an alternative view of microspecies). Species are not evenly distributed between higher taxa; for example a small number of plant genera are unusually species-rich (Cronk 1989; Webb 1991). A good example is Carex in the British Isles (Table 2). To investigate this, the numbers of species in large genera were recorded for each Flora; large genera are defined as those with 20 or more species listed by Stace (1991). It should be noted that one of these genera, Cotoneaster, is composed almost entirely of introduced species. RESULTS AND DISCUSSION Data from 20 regional Floras are shown in Table 1. There are significant correlations (using the non-parametric Spearman rank correlation coefficient) between species richness and genus richness (r,= 0-97, p<0-001) and species richness and family richness (r,= 0-89, p<0-001). Both higher taxon levels investigated are therefore highly correlated with species richness, with variation in genus richness accounting for more of the variation in species richness (r = 0-94) than is accounted for by family richness (r = 0-79). These patterns are found even though the regional Flora data are potentially “noisy”, with different levels of accuracy and detail between different Floras. This suggests that this approach may work with data sets from areas of the world where “noise” is introduced due to the local flora being incompletely known. The results of this study provide an example of a data set with a good positive correlation between higher taxon richness and species richness. Other published examples include relation- ships between family and species richness for: British ferns and British butterflies among 100-km squares, Australian passerine birds using 5° x 5° grid squares and northern and central American bats using grid squares of c. 611,000 km’ (Williams & Gaston 1994). However, Anderson (1995), working on genus richness as a surrogate for species richness in Australian ants, found that “except in limited circumstances, genus richness ... appears to be an unreliable surrogate for species richness in local Australian ant faunas. This may apply more generally to taxa in which relatively small numbers of genera can contribute a large proportion of species.” When the number of species in large genera are compared with total species richness in this British Isles data set a very strong positive correlation was found (r, = 0-91, p< 0-001). This means 272 D. M. WILKINSON that if data were only collected on numbers of species in these 15 genera it would be almost as useful a predictor (r° = 0-83) for comparing areas as a full species list. If such a relationship holds for richer floras which are less well known taxonomically it suggests an option of just recording data from a subset of all genera 1.e., the large ones. In the British Isles data set, removing the large genera has little effect on the relationships between higher taxa as, with the exception of the mainly non-native Cotoneaster, all large genera occur in each of the regional Floras studied. The only approach to validating such heuristic rules is an empirical one, testing these rules at a variety of sites and with a variety of taxa. It is hoped that relationships such as this, tested on the British flora, will also work on tropical forest vegetation. These forests are thought to contain some 40% of the world’s flora (Archibold 1995); mean values of over 230 plant species per 0-1 ha have been described for some of these forests (Crawley 1997). For such areas, comparing lists of higher taxa is a much more realistic possibility than attempting to compile full species lists. The British flora would suggest that comparing sites at the family level could be a good guide to their relative species richness. Comparison at the genus level would provide greater accuracy but would also require more taxonomic work. This study also highlights the importance of regional Floras as biodiversity data sets. Such Floras contain a wealth of data which can be used in many different studies. An attempt to replicate these results using tropical Floras would be of great interest if/when enough data are available. The quotation from Ernst Mayr at the start of this paper suggests that data from plant systematics (and biogeography) can add to our understanding of the world as well as merely describing its variety. It can also contribute to its preservation by providing ways of evaluating the biodiversity of different sites as illustrated by this study. ACKNOWLEDGMENTS I thank Alan Clapham and Daniel Kelly for comments on drafts of this paper. A referee made the interesting suggestion of looking at the larger genera. REFERENCES ALLEN, D. E. (1976). The naturalist in Britain. Allen Lane; London. ANDERSEN, A. N. (1995). Measuring more of biodiversity: genus richness as a surrogate for species richness in Australian Ant faunas. Biological conservation 73: 39-43. ARCHIBOLD, O. W. (1995). Ecology of world vegetation. Chapman & Hall, London. CLapHaM, A. R. (1969). Flora of Derbyshire. Derby Museum, Derby. CRACKLES, F. E. (1990). Flora of the East Riding of Yorkshire. Hull-University Press, Hull. CrAwL_eyY, M. J. (1997). Plant ecology, 2nd ed. Blackwell, Oxford. Cronk, Q. C. B. (1989). Measurement of biological and historical influences on plant classification. Taxon 38: 357-370. Epes, E. S. (1972). Flora of Staffordshire. David & Charles, Newton Abbot. Gaston, K. J. & Spicer, J. I. (1998). Biodiversity, an introduction. Blackwell, Oxford. Gaston, K. J. & Wittiams, P. H. (1993). Mapping the world’s species - the higher taxon approach. Biodiversity letters. 1: 2-8. Goon, R. (1984). A concise Flora of Dorset. Dorset Natural History Society, Dorchester. Gou Lb, S. J. (1996). Dinosaur in a haystack. Jonathan Cape, London. Grana, G. G. (1988). The flora and vegetation of county Durham. Durham Flora Committee, Durham. Hopkinson, I. D. & Casson, D. (1991). A lesser predilection for bugs: Hemiptera (Insecta) diversity in tropical rain forests. Biological journal of the Linnean Society 43: 101-109. Ho.paate, M. W. (1991). Conservation in a world context, in SPELLERBERG, I. F., GOLDSMITH, F. B. & Morris, M. G. The scientific management of temperate communities for conservation, pp. 1-26. Blackwell, Oxford. INGRAM, R. & Nottie, H. J. (1981). The Flora of Angus. Dundee Museums and Art Galleries, Dundee. Janzen, D. H. (1977). What are Dandelions and Aphids? American naturalist 111: 586-589. JANZEN D. H. (1994). Priorities in tropical biology. Trends in ecology and evolution 9: 365-367. Le Sueur, F. (1984). Flora of Jersey. Société Jersiaise, Jersey. May, R. M. (1988). How many species are there on earth? Science 241: 1441-1449. Mayr, E. (1997). This is biology; the science of the living world. Harvard University Press, Cambridge, Massachusetts. : HIGHER-TAXON RICHNESS AND SPECIES RICHNESS 273 NewTon, A. (1971). Flora of Cheshire. Cheshire Community Council, Chester. PANKHUuRST, R. J. & MULLIN, J. M. (1991). Flora of the Outer Hebrides. Natural History Museum, London. Puitp, E. G. (1982). Atlas of the Kent Flora. Kent Field Club, Kent. PriMAVESI, A. L. & Evans, P. A. (1988). Flora of Leicestershire. Leicestershire Museums, Leicestershire. RATCLIFFE, D. A. (1984). Post-Medieval and recent changes in British vegetation: the culmination of human influence. New phytologist 98: 73-100. Roserts, R. H. (1982). The flowering plants and ferns of Anglesey. National Museum of Wales, Cardiff. Rog, R. G. B. (1981). The flora of Somerset. Somerset Archaeological and Natural History Society, Taunton. SaAviIDGE, J. P., Gorpon, V. & Heywoop, V. H. (1963). Travis’s Flora of South Lancashire. Liverpool Botanical Society, Liverpool. Scott, W & Parmer, R. (1987). The flowering plants and ferns of the Shetland Islands. Shetland Times, Lerwick. Simpson, F. W. (1982). Simpson’s Flora of Suffolk. Suffolk Naturalists’ Society, Ipswich. SINKER, C. A. , PACKHAM, J. R. , TRUEMAN, I. C., PERRING, F. H. & PReEstwoop, W.V. (1985). Ecological Flora of the Shropshire region. Shropshire Trust for Nature Conservation, Shrewsbury. Stace, C. A. (1991). New Flora of the British Isles. Cambridge University Press, Cambridge. VANE-WRIGHT, R. I., SmitH, C. R. & Kitcuina, I. J. (1994). Systematic assessment of taxic diversity by summation. Systematics Association special volume 50: 309-326. VincENT, P. (1990). The biogeography of the British Isles. Routledge, London. Wap, A. E., Kay, Q. O. N. & EL.is, R. G. (1994). Flora of Glamorgan. H.M.S.O., London. Wess, D. A. (1991). Presidential address, 1990. Watsonia 18: 247-255. Wess, D. A. & SCANNELL, M. J. P. (1983). Flora of Connemara and the Burren. Cambridge University Press, Cambridge. WILuiAMs, P. H. & Gaston, K. J. (1994). Measuring more of Biodiversity: can higher-taxon richness predict wholescale species richness? Biological conservation 67: 211-217. WIson, E. O. (1992). The diversity of life. Harvard University Press, Cambridge, Massachusetts. Woops, R. G. (1993). Flora of Radnorshire. National Museum of Wales, Cardiff. (Accepted June 1998) oe sis G a i b eure hee bis | tats These: WE) ae we ‘eon ~iehndbes va enn l benim 4 \ seca | (iS rl enaud ea Pia j ihe, ‘uw ‘0 uci ‘van at A bia | ; ney (3 } i" v a | ful rid yee "eet aia Pa pare ’ 1 on" Ha leiden Ph Se ook al ala wind ven: Cheats Lliaidea dane Cetin , ae) 4 vier hen, Bilt tiny sii u ’ Bins ic ’ & ' : itt i hah: Loe ; Watsonia 22: 275—278 (1999) 2S Urtica galeopsifolia Wierzb. ex Opiz (Urticaceae) confirmed for Britain by its chromosome number H. A. McALLISTER School of Biological Sciences, University of Liverpool, Ness Botanic Gardens, Neston, South Wirral, Cheshire, L64 4AY ABSTRACT The ‘stingless’ Fen Nettle from Wicken Fen, Cambridgeshire, England and similar nettles from elsewhere in southern England are found to be diploid (2n = 26) and a new character, non-stinging hair-base diameter of less than 25 um, is given to distinguish this diploid Urtica galeopsifolia Wierzb. ex Opiz from the tetraploid U. dioica L. KEyworbs: Stingless nettle, Urtica dioica, key, Wicken Fen. INTRODUCTION The ‘stingless’ Fen Nettle (Urtica sp.) which grows at Wicken Fen, Cambridgeshire, was thought to be the eastern European U. galeopsifolia, by Geltman (1992). This suggestion stimulated considerable interest and there have since been several reports of U. galeopsifolia at other British locations (Last 1995; Showler 1995; Bull 1995; Killick et al. 1998). Geltman suggested that as U. galeopsifolia was known to be diploid in E. Europe (Geltman 1984), a chromosome count should be made on the Wicken Fen plant. In studies of the Wicken Fen population, Pollard & Briggs (1982) concluded that it was distinct, although other populations had individuals that tended towards the characteristics of the form found at Wicken. They showed that shaded plants tended to have a much lower density of stinging hairs than plants grown in in full sun. However, they found that the density of stinging hairs was genetically based and heritable, and was the best character for distinguishing the Wicken Fen variant. Although Geltman (1984, 1993) records both diploids and tetraploids in U. galeopsifolia and U. dioica, Geltman (1992) regards the former as largely diploid and the latter as largely tetraploid. Given the difficulties of distinguishing between the species on the basis of morphology and the possibility of diploid to tetraploid gene transfer, atypical counts could be the result of misidentification. Accordingly I carried out chromosome counts on a range of Urtica plants to assess the ploidy level and to determine suitable characters for distinguishing between the cytotypes. MATERIALS AND METHODS Living specimens were obtained from Wicken Fen and from three other localities where U. galeopsifolia had been reported (Table 1). These plants were grown in pots and root tips removed for chromosome counting. Counts were made by a modification of the method described by Dyer (1963). The chief modifications to the method were (a) the omission of rinsing following pretreatment and (b) the squashing of the root tips in a mixture of equal parts of lactic and propionic acids. : 276 H. A. McALLISTER TABLE 1 COLLECTION LOCALITIES AND CHROMOSOME COUNTS OF URTICA DIOICA AND U. GALEOPSIFOLIA Species Locality (with grid reference for English localities) Chromosome number Urtica dioica England, Kew, Richmond, Surrey; woodland. J \ == (Ey SY TQ/17.76 Urtica dioica England, Norfolk, Surlingham, footpath (more JV 3 Cy S72 stinging hairs than marsh plant), TG/322.068 Urtica dioica England, Norfolk, North Tuddenham, Dirty Lane, VSO. SP TG/042.148 Urtica dioica England, Wiltshire, Wylye, SU/01.37 NiO, SP Urtica dioica Scotland, Ayr, R. Ayr walk, woodland PA ies (CSV) Urtica dioica Germany, Mecklenburg, Rugen, Wissender 2 =| Cae Klinker, Fagus wood on chalk Urtica dioica Netherlands, Doorn, Gimborne Arboretum, wet 2 = Ce ditchside in shade Urtica galeopsifolia England, Berkshire, Woolhampton, Froud’s Bridge, 2n=¢:26 N. side of river, SU/580.665 Urtica galeopsifolia England, Cambridgeshire, Wicken Fen, TL/55.70 2n\='¢. 26 Urtica galeopsifolia England, Norfolk, Surlingham, Coldham Hall 21 —'C, 261062) Marsh (very wet with Glyceria maxima), TG/324.071 RESULTS Chromosome counts made on nettles collected as U. galeopsifolia or superficially similar morphologically (Table 1) show a clear separation into two cytotypes, U. galeopsifolia as diploid and U. dioica as tetraploid. Observations made in the field confirm that the two cytotypes grow in different habitats, for example at Surlingham the diploid plants were collected from a marsh dominated by Glyceria maxima (Hartman) O. Holmb., while the tetraploid grew in a drier situation close by near a footpath (Bull 1995). The diploid had many fewer stinging hairs than the tetraploid, this being particularly noticeable on the upper surface of the leaf. This observation confirms the view of Pollard & Briggs (1982) that the stinging hair density variation is more or less discontinuous, the Wicken Fen population having a much lower density than any other in their samples. Taken together, these observations suggested that stinging hair density might be the best, if not the only, character to reliably distinguish between the two cytotypes. The morphology of the plants collected in this study showed that all diploids had narrower leaves with many fewer stinging hairs, especially on the upper surface of the leaves, than the tetraploids. It was very difficult to find qualitative characters to distinguish between them though the two cytotypes differ considerably in general appearance and under the microscope. The diploids always look much more delicate and have a much lower density of stinging hairs. The two characters, pubescence density and lowest node of inflorescence, given by Geltman (1993), either did not distinguish between the cytotypes or could be difficult to assess. The pubescence density does not differentiate between the galeopsifolia-like tetraploids and the diploids and the lowest node can be difficult to determine. Known diploids and morphologically similar tetraploids were examined very carefully to see if any other characters could consistently differentiate between them. Under a hand lens or dissecting microscope the indumentum of the tetraploids always seemed to be much coarser than that of the diploids. The difference is due to the greater width of the non stinging hairs in U. dioica giving a much coarser appearance to the indumentum. In diploids the base of these hairs are 20-25 um (0-02—0-025 mm) in diameter whereas in the tetraploids the hair bases are 25-35 um (0-025-0-035 mm) in diameter. The hairs taper gradually from base to apex and are finely patterned with protuberances, but no consistent differences in length or pattern was found. Perhaps the easiest way to identify the diploid would be URTICA GALEOPSIFOLIA 277 by comparison with the common stinging nettle which is usually available. A x 20 hand lens is adequate to observe the difference in hair diameter when comparison is possible. DISCUSSION The detection of two cytotypes which can be separated on morphological grounds means that the diploid growing at Wicken Fen and elsewhere is therefore a distinct biological species. In contrast, superficially morphologically similar specimens from elsewhere are tetraploid (2n = 52). They are not quite so stingless as the Wicken Fen plant but looked rather similar with elongated, narrow, upper leaves. The following key was developed to distinguish between Urtica dioica and U. galeopsifolia. Leaves relatively long and narrow with very few stinging hairs, especially on upper surface of leaf; not flowering till mid July; lowest flowering on 13th to 22nd node; tomentum of non-stinging hairs appearing fine with hair bases 20-25 wm across (detectable by comparison with U. dioica using 20 X hand leMs)..............cccceeeeeeeeeeeees U. galeopsifolia Leaves very variable but with some conspicuous stinging hairs on the upper surface of the leaves; flowering from June; lowest flowering on 7th to 14th node; tomentum of non-stinging hairs appearing coarser with hair bases 25-35 UM ACLOSS ........ cece eeeeeeseeeeeteees U. dioica Most U. dioica plants are of course very distinct and easy to identify but nettles in deep shade and fenlands may greatly resemble U. galeopsifolia and only be identifiable with careful observation. Pollard & Briggs (1992, 1983) concluded that intermediate individuals between the Wicken Fen and ‘ordinary’ nettles occur. This is supported by results here and their possible mode of origin is exemplified by the reported crossing of a Wicken Fen plant with an ‘ordinary’ nettle. Pollard & Briggs (1982) reported six progeny from a single cross between a Wicken Fen and a normal nettle. Of the six offspring they report five varying from intermediate to resembling an ‘ordinary’ stinging nettle. This suggests that hybridization between the ploidy levels occurs readily, producing either triploids or perhaps some tetraploids. Their conclusion of “high interfertility of the Wicken variant with ordinary weedy plants” may be correct but the mechanism needs to be studied now that it is known that two ploidy levels are involved. Hybrids would be expected to be largely triploid with the occasional tetraploid as a result of a non-reductional meiosis in the diploid. Any such tetraploids would be likely to be freely interfertile with tetraploid U. dioica and would result in one way gene transfer from the diploid to the tetraploid (Anamthawatt-Jonsson & Tomasson 1990). Such gene transfer can also occur through triploids (Bielawska 1964). However, because of the ploidy level difference, there is unlikely to be any gene transfer to the diploids, which will therefore remain pure, though gene transfer from tetraploid to diploid is possible and has been documented in Betula (Anamthawatt-Jonsson & Tomasson 1990). The hybrid most similar to U. galeopsifolia would be a tetraploid arising from the union of an unreduced gamete from U. galeopsifolia and a normal reduced gamete from U. dioica. The genome of such a hybrid would have received half of its chromosomes from each parental species. However, repeated interbreeding of such hybrids with U. dioica could result in tetraploid plants more closely resembling U. dioica. It would be very instructive to repeat such diploid U. galeopsifolia x tetraploid U. dioica crosses and examine the ploidy level and fertility of the offspring. The reported intercrossing of Wicken Fen plants yielded only similar progeny, confirming the distinctness and true breeding nature of what we now know to be the diploid. Pollard & Briggs (1982) noted that plants fairly similar in appearance to those at Wicken were found at both Woodwalton Fen, Cambridgeshire, TL/230.840 and South Tawton, N. Devon, SX/655.947 and this suggests other possible sites for the diploid U. galeopsifolia. Dr. J. Edmondson has also drawn my attention to two specimens in the herbarium of the National Museums and Galleries on Merseyside (LIV) which have been identified by Geltman as U. galeopsifolia: v.c. 59, Cheshire: Eastham on the Wirral, collected by J. A. Wheldon in 1894; v.c. H2, North Kerry: Muckross, one mile (1-5 km) N. of Killarney, collected by M. Goodfellow in 1961. 278 H. A. McALLISTER ACKNOWLEDGMENTS I am most grateful to Mr M. Crag-Barber, Mrs B. Last, Dr A Showler and Mr A. Bull for supplying the live material on which this paper is based and to Professor R. H. Marrs and Dr J. Edmondson for much helpful comment on the manuscript. REFERENCES ANAMTHAWAT-JONSSON, K. & Tomasson, T. (1990). Cytogenetics of hybrid introgression in Icelandic birch. Hereditas 112: 65-70. BieLawska, H. (1964). Cytogenetic relationships between lowland and montane species of Campanula rotundifolia L. group. C. cochleariifolia Lam. and C. rotundifolia L.. Acta Societatis Botanicorum Poloniae 33: 15-44. BuLL, A. (1995). Urtica galeopsifolia in Norfolk. B.S.B.I. news 69: 30-31. Dyer, A. F. (1963). The use of lacto-propioinic orcein in rapid squash methods for chromosome preparations. Stain technology 38: 85-90. GELTMAN, D. V. (1984). Cytotaxonomic studies of the species of the genus Urtica (Urticaceae) in the flora of the USSR. Boanicheskii zhurnal 69: 1524-1530. GELTMAN, D. V. (1992). Urtica galeopsifolia Wierzb. ex Opiz (Urticaceae) in Wicken Fen (E. England). Watsonia 19: 127-129. GELTMAN, D. V. (1993). Urtica L. in Tutin, T. G. et al., eds Flora Europaea 1: 79-80, (2nd ed.). Cambridge University Press, Cambridge. KILLICK, J., PERRY, R. & WOoDELL, S. (1998). The Flora of Oxfordshire. Pisces Publications, Newbury. Last, B. (1995). Another site for the non-stinging nettle. B.S.B.J. news 68: 10. POLLARD, A. J. & Bricas, D. (1982). Genecological studies of Urtica dioica L. I. The nature of intraspecific variation in U. dioica. New phytologist 92: 453-470. PottarbD, A. J. & Briccs, D. (1984). Genecological studies of Urtica dioica L. I. Patterns of variation in Wicken Fen, Cambridgeshire, England. New phytologist 96: 483-499. SHOWLER, A. (1995). Fen nettle (Urtica galeopsifolia) in Berkshire. B.S.B.I. news 69: 31. (Accepted July 1998) Watsonia 22: 279-281 (1999) 279 Lysimachia punctata L. and L. verticillaris Sprengel (Primulaceae) naturalised in the British Isles H. A. McALLISTER School of Biological Sciences, University of Liverpool, Ness Botanic Gardens, Neston, South Wirral, Cheshire, L64 4AY ABSTRACT Four species of Lysimachia, L. punctata and L. verticillaris, and to a lesser degree L. ciliata and L. vulgaris, are much confused in gardens and when naturalised. L. punctata is easily distinguished from L. verticillaris by the absence of an orange area at the petal base, the leafy nature of all bracts, the leafy continuation of the inflorescence axis beyond the inflorescence, the more elongate rhizomatous habit and, in Britain, the absence of viable seed. These distinguishing characters seem not to have been previously noticed. L. vulgaris is native but is also found as a garden escape. The other three species are found occasionally as escapes from cultivation though L. verticillaris has not previously been recognised as distinct from L. punctata. L. punctata and L. verticillaris have chromosome numbers of 2n = 30 while L. ciliata has 2n = c. 96. KeEyworps: yellow loosestrife, chromosome number. INTRODUCTION This note was prompted by difficulties encountered in identifying yellow loosestrifes (Lysimachia spp.) both in cultivation and naturalised in the British Isles. I was very familiar with L. punctata L. naturalised and as a garden plant in Argyll where it is often found in ditches near abandoned crofts. However, at Ness Botanic Gardens on the Wirral a quite different plant was labelled L. punctata. The orange spot at the petal base, branched inflorescence branches, absence of continuation of the leafy axis beyond the inflorescence and the fact that it produced seed, easily distinguished this species from what I knew as L. punctata - which I had never seen to produce seed and therefore assumed to be self-incompatible. There were, therefore, two quite distinct taxa commonly being grown under the same name. Originally I identified the plant in cultivation at Ness as L. ciliata L. because of the presence of the orange spot at the petal base (Ferguson 1972), but this identification was questioned by a visitor who provided a living specimen of L. ciliata which was clearly CSL from either of the two species already in cultivation and was named correctly. There was clearly confusion in the literature. Huxley (1992) mentions L. verticillata (sic) as being very similar to L. punctata. Ferguson (1972) regards L. verticillaris Sprengel as synonymous with L. punctata, but Leblebici (1978) describes L. punctata and L. verticillaris as distinct, distinguishing them primarily on petiole length. Accordingly a morphological study has been made of the upright Lysimachia species likely to be encountered native or naturalised in the British Isles and the following key devised, a summary of distinguishing characters being given in Table 1. DISCUSSION L. vulgaris is distinct with its terminal, paniculate inflorescence, L. ciliata equally so with its flat flowers on long pedicels and its ciliate petioles and L. terrestris (L.) Britton with its smaller spotted flowers and bulbils. The other two species are much confused and usually grown under the name L. punctata. However, they are easily distinguished by the characters mentioned in Table 1. No previous mention seems to have been made of the orange area at the petal base in L. verticillaris, its 280 H. A. McALLISTER TABLE 1. DIFFERENTIAL CHARACTERS AMONG FOUR UPRIGHT SPECIES OF LYSIMACHIA L. vulgaris L. ciliata L. punctata L. verticillaris Petal base pale orange pale orange Bracts at extreme tip of leafy leafy very leafy subulate or leafy inflorescence Bracts in upper inflores- subulate leafy, ovate leafy, ovate subulate cence Inflorescence panicle panicle-raceme raceme raceme-panicle Axillary branching in raceme - single flower or 2 flowers on 2 flowers or inflorescence single flower reduced raceme ebracteolate reduced raceme with bracteoles pedicels with bracteoles Flower shape campanulate flat campanulate campanulate Pedicel length <2:5cm >2-Scm <2-5cm <2-S5cm absence in L. punctata, or the fact that the inflorescence is usually terminal in L. verticillaris, — lacking the leafy apical rosette to the racemose inflorescence which is always so conspicuous in L. punctata. All the bracts in L. punctata are leafy whereas only the lower and very rarely the extreme uppermost are leafy in L. verticillaris; the bracts in most of the upper part of the inflorescence are subulate and inconspicuous. This results in a much less leafy appearance to the inflorescence of L. verticillaris. On vigorous stems both species can produce lateral inflorescences similar to the primary one. L. verticillaris always produces axillary reduced racemes in the axils of bracts on the lower part of the raceme whereas bracts of L. punctata always bear paired flowers in their axils. These flowers are borne on ebracteate pedicels though the pedicels are often fused for part of their length. The inflorescences of L. verticillaris may therefore sometimes look almost paniculate or be a panicle of racemes. Another very evident difference in cultivation is that L. verticillaris forms clumps of stems which do not spread laterally to any great extent while L. punctata is strongly rhizomatous and forms dense, extensive patches. The most obvious distinguishing character in living flowering specimens is the orange spot at the petal base in L. verticillaris. This character is much less obvious in herbarium specimens. However, some of the other characters mentioned, perhaps especially the nodal red coloration in L. verticillaris, may not be present in all specimens in the wild, only in the material introduced to cultivation. Many of the other distinguishing characters mentioned above are also much less evident on herbarium specimens though very obvious when living material is studied. Difficulties of identification may also be encountered, with both herbarium and living material, because of the state of maturity of the specimens. In immature specimens, or those taken early in the flowering period, the state of the bracts in the upper parts of the inflorescence will not be evident. Supplementary characters given both in the key and table should, however, allow the correct identification of such specimens. Leblebici (1978) describes the two species as having non-overlapping ranges, L. punctata occurring in Europe and Western Turkey while L. verticillaris is found in north and east Turkey, Caucasia, Crimea and N. Iran. L. verticillaris always produces liberal quantities of viable seed whereas L. punctata has never been seen to produce any seed, even when grown in close proximity to L. verticillaris. Therefore it seems that L. verticillaris is self-compatible while L. punctata is self-incompatible and that the two species do not normally interbreed. This lack of interbreeding supports the retention of the two taxa as distinct species. Both L. verticillaris and L. punctata have chromosome numbers of 2n = 30 while L. ciliata has 2n = c. 96 (counts made in the course of this study). L. vulgaris is recorded as having 2n = 56, 84 (Ferguson 1972). HABITAT REQUIREMENTS It is interesting that Leblebici (1978) describes L. punctata and L. verticillaris as occupying wet LYSIMACHIA PUNCTATA AND L. VERTICILLARIS 281 habitats in the wild and it is nearly always in wet situations that they are found naturalised. In gardens they are usually grown in herbaceous borders where they grow more or less satisfactorily as long as they are free from competition. At Tighnabruaich, Argyll in Scotland naturalised populations of L. verticillaris occur between the road and the shore (Grid ref. NR/990.738) alongside naturalised plants of Persicaria campanulata (Hook. f.) Ronse Decraene (Polygonum campanulatum Hook. f.), Geranium x oxonianum Yeo, and Rumex pseudoalpinus Hoefft, all species of wet habitats in their native ranges and when naturalised in Britain often found growing in flushes. Voucher specimens of L. verticillaris are deposited in LIV, E and BM. This is the first record of the species for the British Isles; it is not mentioned in Clement & Foster (1994) but many records of L. punctata are likely to be L. verticillaris. KEY TO FOUR CONFUSED SPECIES OF LYSIMACHIA FOUND IN THE BRITISH ISLES la. Petiole ciliate; leaves glabrous, mostly opposite; pedicels more than 2-5 cm; open flower ibe RNR EE SE Ne ean gaan Aud sued a neaet soled suaus sockakobissade vou dewateadsensebanedidaseberensd ciliata Ib. _Petiole not ciliate; leaves pubescent, mostly verticillate; pedicels less than 2:5 cm; open IGS TAC ANP AMULALC aeeee te eeeee ney ok Micah sce sasaameeNGetccsdeiesesstnesducecsestencesacadceetdulocerss tebseseoreaosuteuts 2 2a. Inflorescence paniculate; flowers borne singly in axils of bracts; stem leaves usually in WHOIS OF (LESS scssagdhooadcocdosogsaccesucbdece soonaaeReey CONCH? SNR 7 ia Heo aRMMME = HHnE AF Stes sD Hiae Aes man RE vulgaris 2b. Inflorescence racemose or a panicle of racemes; flowers paired in axils of bracts; stem leaves usually in whorls of more than three .0.......... ccc ceecccssssecessscceessescssssseeensessseeseseeeeseees 3 3a. Petals with orange flush at base; some axillary branches in inflorescence themselves - branched and bearing several flowers subtended by bracteoles; some bracts subulate and shorter than or equal to pedicel; petiole more than 5 mm; nodes and leaf bases purplish; infloresence racemose to a pamicle Of raCEMES ............eeeeseecccesssseeeeeseeeeseeeseees verticillaris 3b. Petals paler towards base; axillary branches in inflorescence single or 2-flowered but pedicels lacking bracteoles; all bracts leafy and much longer than pedicel; petiole less than ACKNOWLEDGMENTS I am grateful to Professor R. H. Marrs, M. Sanford and anonymous reviewers for helpful comments on the manuscript. REFERENCES CLEMENT, E. J. & Foster, M. C. (1994). Alien plants of the British Isles. Botanical Society of the British Isles, London. Fercuson, L. F. (1972). Lysimachia L. in Tutin, T. G. et al. eds. Flora Europaea 3: 26-27. Cambridge University Press, Cambridge. Hux.ey, A. et al., eds. (1992). The new Royal Horticultural Society Dictionary of gardening. Royal Horticultural Society, London. LeBLEBICI, E. (1978). Lysimachia L. in Davis, P. H. ed. Flora of Turkey 6: 135-138. Edinburgh University Press, Edinburgh. _ (Accepted April 1998) Ree) ae ee ARAL OVER af Jovileveidn’ ide oun sap i yihorod falter peal 0 nen wot, Yo Boviteunen’ baaltowe” mt ‘hoe AGES, O Sei oe et onary sith “te "Wie at unis ihe: Shana wh vA bale, sitwort, t oY oie mist at ih pa Boyrpomngeatt othareedt JMh win th wa si titcige a6 « alinhewealag ho ebveten yrienT ied RAND EY HOE % Meh ct ei baton fot ele yotal dete at baie 4 mad Dee an) ore y AAG: by 43 rae ates abt had m4 . " i nee i, oat mo, aig, . > client. ey swat! oq. | a OE nists. BB igh a 3 Peery?) Cea et aA ‘4 eduels ae eee yey so fis Vue bl a dion Pelee ioe ‘ sere een Ree teed Beal elie wins fi sy ov Hypoen jeenuadsg ayaa | o, BA ae es ee a fale oe ee sf i prnciniat Lom sos li Ai WEE RUAN tee giao! mele | i agers BA Bieta: it Weel ie ua swan at DENSMEENY os Socios 5 chooks a de iia oe aon AP nage ARE water nde. hohe LHe, 21a. (EATS am he 4 slaty rH * A “ cae en at 16 + + ou) Wan Sal) uae nti Or Batkiis ic Re Onoe rodent! i AE AAY oavonh separa +) HAW rd yd babysit tai isioveie Pies’ aii: Wace Tas hE Gay here nil Syne We yey Pexboy Oe he Heit Ay TS a | ‘na LRG aN Sg it Caneahhd eh ae TS whe: her Sarin ee tC veRtinen ‘ite iit’ abs . (acy eet SEG TV oSe a HOO OU Lr eat ‘ape eS ete , f t q 4 bn ° rt by ea 7 | . 7 “ ’ a? : 4 y hay h rs, i oes geshck BF ASL Oe a TAREE Aen oh . y Hi ; ’ } 1 P G iG oe vigeenon Hoelod 4! weed ooo bile pad 0b ae ae, . . ‘ ‘are? Sag YAS by Da Macht: | a cal ow Gln ches + (ey ids AP eae RG hohe Bisledoe? aie eva SR isda ish ae geet ta ie A a iy if ‘*) ; iv é Pxvivpie Nes ants te _ pass iy ira hoa A by me NN fat 4A ha hess Mtr Wee ba eee ne ey is Paes el ae ‘ y sy rE, 4 ae a one a nhs . (oe 4 y 4 ade ae ok uP aie ‘i dicen, 0 Aisin: Prise in da. ine < = é pleas ated See - er — 2 ROME, sinngy eu M vroieill le a mt re i i, 7 yet TA «ete Ney ih ah ou hip eee ee ee a ea appa HAL hl 4 ee Ane Med, aoa Wi ee oi Oth pie ie ne eal dar Vein) onan | ' i i iahgenlieele er ws hat H AY ‘ BOTANICAL SOCIETY OF THE BRITISH ISLES aoa) ee The B.S.B.I. was founded in 1836 and has a membership of 2,850. It is the major source of information on the status and distribution of British and Irish flowering plants and ferns. This information, which is gathered through a network of county recorders, is vital for the conservation of the plants and is the basis of the Red data books for vascular plants in Great Britain and Ireland. The Society arranges conferences and field meetings throughout the British Isles and, occasionally, abroad. It organises plant distribution surveys and publishes plant atlases and handbooks on difficult groups such as sedges and willows. It has a panel of referees available to members to name problem plants. Through its Conservation Committee it plays an active part in the protection of our threatened plants. It welcomes all botanists, professional and amateur alike, as members. Details of membership and any other information about the Society may be obtained from: The Hon. General Secretary, Botanical Society of the British Isles, c/o Department of Botany, The Natural History Museum, Cromwell Road, London, 5W7 5BD. INSTRUCTIONS TO CONTRIBUTORS Scope. Authors are invited to submit Papers and Notes concerning British and Irish vascular plants, their taxonomy, biosystematics, ecology, distribution and conservation, as well as topics of a more general or historical nature. Authors should consult the Hon. Receiving Editor for advice on suitability or any other matter relating to submission of manuscripts. Papers and Notes must be submitted in duplicate, typewritten on one side of the paper, with wide margins and double-spaced throughout. Submission of final edited copy on computer disc will be requested, but by two hard copies of the text are acceptable if computer facilities are not available. Format should follow that used in recent issues of Watsonia. Underline where italics are required. Names of periodicals should be given in full, and herbaria abbreviated as in British and Irish herbaria (Kent & Allen 1984). The Latin names and English names of plants should follow the New Flora of the British Isles (Stace 1997). Further details on format can be obtained from the Hon. Receiving Editor or by viewing the website at: http://members.aol.com/bsbipubs/ watsonia.htm. Tables, figure legends & appendices should be typed on separate sheets and attached at the end of the typescript. Figures should be drawn in black ink or be laser-printed and identified in pencil on the back with their number and the author’s name. They should be no more than three times final size, bearing in mind they will normally be reduced to occupy the full width of a page. Scale-bars are essential on plant illustrations and maps. Lettering should be of high-quality and may be done in pencil and left to the printer. Black and white photographs can be accepted if they assist in the understanding of the article. Contributors must sign a copyright declaration prior to publication which assigns the copyright of their material to the Botanical Society of the British Isles. Twenty-five offprints are given free to authors of Papers, Notes and Obituaries; further copies may be purchased in multiples of 25 at the current price. The Society takes no responsibility for the views expressed by authors of Papers, Notes, Book Reviews or Obituaries. Submission of manuscripts Papers and Notes: Mr M. N. Sanford, c/o The Museum, High Street, Ipswich, Suffolk, IP] 3QH. Books for Review: Mr D. A. Pearman, The Old Rectory, Frome St Quintin, Dorchester, Dorset DT2 OHF Plant Records: the appropriate vice-county recorder, who should then send them to Dr C. D Preston, Biological Records Centre, Monks Wood, Abbots Ripton, Huntingdon, PEI7 21S. Obituaries: Mrs M. Briggs, 9 Arun Prospect, Pulborough, West Sussex, RH20 IAL. Back issues of Watsonia are handled by Dawson UK Limited, Cannon House, Folkestone, Kent CT19 5EE to whom orders for all issues prior to Volume 21 part 1 should be sent. Recent issues (Vol 21 part 1 onwards) are available from Mr M. Walpole, B.S.B.L, 68 Outwoods Road, Loughborough, Leicestershire, LE11 3LY. Watsonia February 1999 Volume twenty two Part three Contents Swan, G. A. Identification, distribution and a new nothosubspecies of Trichophorum cespitosum (L.) Hartman (Cyperaceae) in the British Isles and N. W. Europe HOLLINGSworTH, P. M. & Swan, G. A. Genetic differentiation and hybridisation among subspecies of Deergrass ae cespitosum (L.) Hartman) in Northumberland ae Aksoy, A., Hate, W. H. G. & Dixon, J. M. Towards a simplified taxonomy of Capsella bursa-pastoris (L.) Medik. (Brassicaceae) Se eee Ricu, T. C. G. Conservation of Britain’s diversity: F’ es lutescens Jordan (Asteraceae), Red-tipped cudweed DeEARNLEY, T. C. & Duckett, J. G. Juniper in the Lake District National Park. A review of condition and regeneration Witkinson, D. M. Correlations between meee richness and a richness in the British Isles flora ... McA.uisterR, H. A. Urtica galeopsifolia Wierzb. ex ian (Urticaceae) confirmed for Britain by its chromosome number MCcALuIsTER, H. A. Lysimachia punctata L. and L. verticillaris Sprengel (Primulaceae) naturalised in the British Isles ... Notes Lansdown, R. V. A terrestrial form of Callitriche truncata Guss. subsp. occidentalis (Rouy) Braun-Blanquet (Callitrichaceae) | Book REVIEWS OBITUARIES REPORT Published by the Botanical Society of the British Isles ISSN 0043-1532 Typeset by D. K. & M. N. SANFORD Printed in Great Britain by THE BOOK COMPANY, PO BOX 243, IPSWICH, SUFFOLK 209-233 235-242 243-250 25 1-260 261-267 269-273 275-278 . 279-281 283-286 287-291 293-295 297-300 A mt Nel Bi Journal of the tanical Society of the British Isles 1e 22 Part 4 August 1999 Briggs, J. R. Edmondson, D. impson L. Kelly S D. A. J : >. Rushton, M. N. sanford Botanical Society of the British Isles Patron: Her Majesty Queen Elizabeth the Queen Mother Applications for membership should be addressed to the Hon. General Secretary, c/o Department of Botany, The Natural History Museum, Cromwell Road, London, SW7 5BD, from whom copies of the Society’s Prospectus may be obtained. Officers for 1999-2000 President, Mrs M. Briggs President elect, Dr G. Halliday Vice-Presidents, Dr R. J. Gornall, Mr M. Walpole, Mr D. J. McCosh Honorary General Secretary, Mr R. G. Ellis Honorary Treasurer, Mr M. E. Braithwaite Editors of Watsonia Papers and Notes, D. L. Kelly, M. N. Sanford*, D. A. Simpson Plant Records, C. D. Preston Book Reviews, D. A. Pearman Obituaries, M. Briggs *Receiving editor, to whom all MSS should be sent (see inside back cover). © 1999 Botanical Society of the British Isles The Society takes no responsibility for the views expressed by authors of Papers, Notes, Book Reviews or Obituaries. Accredited with the International Association for Plant Taxonomy for the purpose of registration of new non-fungal plant names. The cover illustration of Euphorbia pe: L. (irish Spurge) was drawn by Rosemary Wise. Watsonia 22: 301—315 (1999) 301 Biology, genetic variation and conservation of Luronium natans (L.) Raf. in Britain and Ireland Q. O. N. KAY, R. F. JOHN School of Biological Sciences, University of Wales, Swansea SA2 8PP and R. A. JONES Countryside Council for Wales, Ladywell House, Newtown, Powys SY16 1RD ABSTRACT Luronium natans (Floating Water-plantain) is a European endemic aquatic plant which is now rare and threatened across most of its extant range. Confusion with similar species has led to misunderstanding of its distribution and status in Britain and Ireland, but oligotrophic upland lakes now seem to hold the main populations. Isozyme studies show comparatively high levels of genetic variation among native populations in Wales, contrasting with the lack of variation reported in some closely-related aquatic species, but there is close similarity among samples of plants from canal populations on the Welsh borders. It has spread into this canal habitat relatively recently (post-1850), and the isozyme patterns found there indicate that the canal populations have originated from a native lake population connected to the canals by feeder streams. The greatest range of genetic variation now appears to remain in a few native “core” sites, and the survival of these populations is thus of particular importance. Conservation priorities are discussed in relation to the biology, distribution and metapopulation structure of the species in Britain and Ireland. Keyworps: Floating Water-plantain, aquatic plants, isozyme variation, Wales, threatened species, oligotrophic lakes. INTRODUCTION Luronium natans (L.) Raf. (Alismataceae) is a stoloniferous aquatic perennial which grows in a range of habitats, and is phenotypically very plastic. The strikingly different forms that it can assume in different habitats nearly all resemble commoner aquatic species, with the result that it has often been overlooked or confused with these commoner plants (Ferguson, Briggs & Willby 1998). In shallow water L. natans produces stalked leaves with ovate or elliptical blades, up to about 4 x 1-4 cm, sometimes larger, which superficially resemble the floating Icaves of some common Potamogeton species. When growing on exposed mud it produces similar but smaller ovate leaves with stiffer, shorter stalks, and then closely resembles Baldellia ranunculoides, even when it is in flower. In deeper water (0-7—2 m, perhaps up to 4 m in particularly clear lakes) it grows as “isoetoid” rosettes of linear-triangular leaves, about 0-5 cm wide at the base and 5-7 cm long, without any expanded lamina. These rosettes are practically identical to juvenile Alisma plantago-aquatica, and very similar when seen at a distance to those of Jsoetes spp. and Littorella uniflora, both of which are often abundant in habitats that are suitable for Luronium natans, and may not be recorded in shore-based sampling. Finally, in flowing water it can produce long (50-60 cm), strap-like, flexible leaves about 0-5—0-7 cm wide, in dense masses, resembling the leaves of the species of Sparganium that grow in similar habitats. The forms (illustrated by Jones & Rich (1998)) appear to be phenotypically freely interchangeable, both in cultivation and in nature, with leaves of different forms being produced on one and the same rosette under appropriate conditions. Their common feature, found in otherwise similar aquatic species only in Baldellia ranunculoides, is the combination of slender stolons, 5 cm or more in length, with laminar leaves produced in 302 Q. O. N. KAY, R. F. JOHN AND R. A. JONES rosettes. The stoloniferous forms of Baldellia ranunculoides that most closely resemble L. natans can be distinguished in the field by the characteristic smell of coriander in their crushed foliage. In Bnitain and Ireland Luronium natans is known from three main habitats. Its chief natural habitat is now in oligotrophic, moderately acidic lakes and pools, in upland areas but generally not at high altitudes (up to about 400 m). In these sites it can occasionally be detected by the presence in shallow water of flowering plants with floating leaves, but more often remains inconspicuous and unsuspected as submerged isoetoid rosettes, vegetative and growing at depths of 1-2-5 m. It appears to have been lost from similar, formerly oligotrophic lakes in the lowlands, for example the Shropshire meres and the larger pools and lakes in Anglescy (Ynys MO6n) as a consequence of eutrophication, mainly during the early part of the century. A second characteristic habitat, which sull survives in a few sites, is in shallow oligotrophic or mesotrophic ponds or pools on grazed heathland or commonland in low-lying western maritime areas. Scattered references to its former occurrence in such sites (e.g. Davies 1813) suggest that L. natans may have had a fairly wide distribution in ponds of this type in the past, but has drastically declined as they have been lost or altered by drainage, eutrophication, changes in agricultural practice or overgrowth. L. natans 1s now known to grow at only two such sites in western Britain, on Ramsey Island (Ynys Dewi) and Dowrog Common, both in south-western Wales. Heathland ponds of this type are or were often man-made, with their suitability as a habitat for L. natans maintained or enhanced by human or livestock activity, and at least some of the populations of L. natans in such sites can be seen as satellite populations utilising relatively transient and essentially artificial sites. Similar small and probably transient populations found in isolated farm ponds in Pembrokeshire, in scattered sites across the English midlands, and (perhaps introduced) in Norfolk (M. Wade pers. comm., Driscoll 1992) reflect this more opportunistic aspect of the ecology of L. natans, as does its third presently characteristic habitat, which is in abandoned or little-used canals. It has spread into this habitat fairly recently (since about 1850), colonising canals on the Welsh border and in north-central England, sometimes becoming locally abundant (Willby & Eaton 1993) but subsequently declining in at least some sites (Briggs 1988) as a consequence of eutrophication, overgrowth, increased boat traffic and other possible factors. It seems possible that the canal habitat effectively reproduces a formerly widespread but now very rare natural habitat of L. natans, in clear, oligotrophic or moderately mesotrophic, slow-flowing rivers. In Britain L. natans has been recorded from seven such river sites, but it is now known from only two, both in Wales. As with lowland pool and lake sites, suitable river habitats would almost certainly have been more common in the past, before widespread channel-straightening, eutrophication and sediment deposition took place, and river populations of L. natans have most probably declined as a result (Davies 1813; Lockton and Whild 1997). Whilst the present canal populations of L. natans clearly grow in an artificial habitat, their relationship to other populations of the species has been uncertain. The interconnection between and superficial similarity of canals to sluggish river channels suggests that canal populations might have had river populations as their source, although it is also possible that nver populations are themselves satellites of “core” populations in upland Wales. A number of the current and former sites for L. natans in rivers occur downstream of lakes containing large, established populations. Luronium natans is a European endemic species, with a distribution of the Suboceanic Temperate type (Preston & Hill 1997). Its range extends from north-western Spain and Ireland in the west to Lithuania, Bulgaria and Moldavia in the east. On much of the European continent it is now very rare and declining, with a scattered and disjunct distribution, and many surviving populations are reported to be under threat from drainage, eutrophication or acidification of their freshwater habitat (Roelofs 1983; Hanspach & Krausch 1987; Fritz 1989; Ferguson 1991; Rodriguez-Oubina & Ortiz 1991; Willby & Eaton 1993). As a result the species has been listed on the Berne Convention Appendix 1 (which requires signatory states to prohibit taking, and to take measures to conserve listed species), and the EC Habitats Directive Annexes IIb & IVb (which require designation of protected areas and special protection measures for them). Consequently, L. natans is now listed on Schedule 8 of the United Kingdom Wildlife and Countryside Act (revised 1992), which makes it an offence to pick, uproot, sell or destroy the species, and is included among the “Short List” species of the UK Biodiversity Action Plan, with a commitment to maintain and, if possible, enhance its present range. In the British Isles its distribution has been rather poorly understood in the past, in part because of confusion with other commoner species, but it now seems clear that L. natans has a stronghold LURONIUM NATANS IN BRITAIN AND IRELAND 303 in Wales, where it is fairly widespread in the uplands (Ellis 1983; Preston 1994). With the decline of heathland ponds, native or long-established lowland populations in adjacent English counties have dwindled greatly and may recently have become extinct in this habitat (S. Whild, pers.comm.). Past reports of the occurrence of L. natans in the mountainous Lake District of north-western England (Cumbria) were repeatedly copied and cited in County Floras and other publications (Turner & Dillwyn 1805; Smith 1828; Hooker 1831; Baker 1885; Hodgson 1898; Wilson 1938; Perring & Walters 1976; Halliday 1978; Willby & Eaton 1993), but investigation of these records, all dating from the nineteenth century, showed that they were questionable and in one case the result of a copyist’s error (Kay & John 1995; Halliday 1997), and as such they were omitted from the maps published by Preston (1994) and Preston & Croft (1997). Subsequent re-investigation of likely habitats in Cumbria (G. Halliday, pers. comm.) has however shown that L. natans is locally abundant in at least two of the larger lakes, Derwentwater (Halliday 1997) and Bassenthwaite Lake, and it has also been found in a pool near Ullswater (G. Halliday, pers. comm.), so in this light it now seems probable that Greville’s early record of the species from Derwentwater (Turner & Dillwyn 1805), which was first called into question by Hodgson (1898) was in fact correct. Furthermore, there is now evidence that L. natans has been even more widely overlooked. Nineteenth-century records of its occurrence in south-western Scotland and western Ireland (e.g. Hooker 1831) had, in the absence of more recent records, come to be regarded as probable errors. However, Rich, Kay & Kirschner (1995) have recently found a new locality for the species in Ireland, and they have shown that several older records from other sites in Ireland are certainly or probably correct. It now seems likely that L. natans, although scarce, may be fairly widely distributed in western Ireland, at least from Killarney (where it was last seen in 1886) through Clare (1882) to Galway (1994). An equally interesting series of new records of L. natans has recently been reported from central and western Scotland; here, lake surveys have shown that it occurs at several sites in Argyll, although its status there is uncertain (N. Willby, pers. comm.). These records from Ireland and Scotland, and its rediscovery in the English Lake District, show that L. natans is considerably more widespread in western Bnitain and Ireland than had been thought, and suggest that it may be present but undetected in other sites. The aims of the present study, which commenced in 1993, were to characterise the reproductive biology of L. natans, to assess its patterns of genetic variation using isozyme analysis, to interpret these findings in terms of its population and metapopulation structure, and to consider their implications for its conservation. MATERIALS AND METHODS Observations on reproductive biology were made both in the field and in cultivation. Plant samples for isozyme analysis were collected (under licence) by hand or grab sampling at intervals of at least 2 min pond and stream populations and 5—20 m in lake populations. Grapnels do not attach well to L. natans, and simple diving equipment — a face-mask, snorkel and wet-suit — is the most effective means of sampling colonies below 1 m depth. One or more clones from most populations were grown on for further study, without difficulty, rooted in fine gravel in a 15 cm depth of rainwater in open-topped polythene containers 12—18 cm in diameter on a part-shaded open-air hardstanding. Isozyme analysis was carried out by horizontal starch gel electrophoresis, using about 0-5-1 cm’ of fresh leaf material from each sample. The electrophoresis procedure was similar to that described by Shields, Orton & Stuber (1983) and Lack & Kay (1986). Staining was carried out using recipes following Shaw & Prasad (1970). RESULTS REPRODUCTIVE BIOLOGY Luronium natans flowers during July and August, and appears to be adapted for both self- pollination and insect visitors. Plants growing in less than about 60 cm of water, or on exposed mud, produce bowl-shaped flowers which open to about 15 mm in diameter in sunny and calm weather, rising to or above the surface on long pedicels if the plant is submerged. In cultivation, the 304 Q. O. N. KAY, R. F. JOHN AND R. A. JONES TABLE 1. LURONIUM NATANS — SITES, SAMPLE SIZES, NUMBERS OF ISOZYME GENOTYPES AND NUMBERS OF VARIABLE LOCI FOUND Population code, site name and grid reference No. of No. of isozyme No. of samples collected § genotypes found variable loci WALES . Llyn Cwellyn SH/565.548 . Llyn-y-Dywarchen SH/560.534 . Llyn Tegid SH/88.35 . Llangollen Feeder Canal SJ/210.426 . Montgomery Canal, Four Crosses SJ/258.191 . Montgomery Canal, Nag’s Head SO/194.990 . Montgomery Canal, Fron SO/167.965 . Llyn Hir SN/789.677 9. Llyn Teifi (all) SN/78.67 9A. Llyn Teifi A (South Bay) SN/785.675 9B. Llyn Teifi B (South Shore) SN/784.675 10. Llyn Egnant SN/792.675 11. Afon Teifi, Cors Caron SN/69.64 12. Afon Teifi above ‘Flash’ SN/675.620 13. Afon Teifi below ‘Flash’ SN/677.640 14. Llyn Eiddwen SN/606.670 15. Llyn Fanod SN/603.644 16. Ramsey Island, West Pond SM/702.235 IRELAND 17. Invermore Lough (Connemara) L/899.390 CONNNBWNY —a ee KH WWNYNwWWH KF KNNNNNYN SND OE Ooo S Swe L Sos SOOWNHKENNNWNWHK KH WH Ho — i=) flowers last for only a day, but regularly attract small flies. Their three delicate white petals have a conspicuous yellow base, and are unscented. Their nectar provision, if any, was not quantified, but they resemble the nectar-providing flowers of Ranunculus subgenus Batrachium (water crowfoots) which grow in similar habitats. In windy conditions, or when produced on plants growing at depths greater than about 60 cm, when the flowers fail to reach the surface, the petals do not open and self-pollination takes place cleistogamously within the closed flower. G. Halliday (pers. comm.) has observed the production of several long-stalked cleistogamous flowers from each of many ascending stoloniferous stems in plants growing in about 2 m of water in Derwentwater. There are normally six stamens per flower, and up to 13 carpels (an average number of 12-5 was found in samples from Ramsey Island (Ynys Dew1)), each containing a single ovule, and forming an achene when fertilized. The pollen grains are morphologically distinctive, bluntly octagonal in cross- section; the pollen fertility of a sample from Ramsey, assessed as acetocarmine stainability, was 83-97%, with about 12000 pollen grains per flower. The pollen:ovule ratio was 973, suggesting adaptation for at least partial outbreeding. No formal test of compatibility relationships was made, but the production of seeds by submerged and apparently cleistogamous flowers indicates self- compatibility. In suitable conditions, for example on the exposed mud of Llyn Tcifi (Chater 1990) or in sheltered parts of Llyn Glaslyn (Catherine Duigan, pers. comm.), and in the Montgomery and Rochdale Canals (N. Willby, pers. comm.) flowering plants can produce conspicuous displays. It seems probable that allogamy is possible only in plants that are growing in fairly shallow water where flowering shoots can extend to the surface, or on exposed mud, and that seed production is most abundant under these conditions. The small (1-2 mm) cylindrical achenes have no special adaptations for dispersal, and in water sink when released from the receptacle, although seedlings float (Ridley 1930) and it seems likely that the seeds can be dispersed by waterfowl. Seedling establishment is sometimes conspicuous (as, for instance, along the strand-line on Llyn Teifi, Cardiganshire), but the characteristically vigorous stoloniferous spread shown by L. natans suggests that reproduction within individual lakes, rivers or canal systems is likely to be predominantly clonal, by vegetative spread of stolons and detached rosettes. Rosettes and stolons might also occasionally become attached to the legs or necks of waterfowl, enabling vegetative 305 LURONIUM NATANS IN BRITAIN AND IRELAND 000:1 000-1 L91-0 91 000:1 L9L‘0 eet 0 tee -0 L91-0 Sl l (6 £E ESICON BEes:0 00S:0 00S°0 tte-0 00S:0 tttO ettt0 E [ [ ett-0 ett0 O00S:0 00:0 tte0 O00S0 00:0 c l [ 000! OSZ0 00S-0 0Sc:0 00-0 l E G L910 =6L91-0 =—L91-0 CEC O It- OF e290 CEEO SUNOF BZ9T-0 ett0 LILO CEE0 SITVILO® 291-0 L910 =6L9T-:0 ~—s L910 vi el Cl l (3 ttt0 ccc-0 111-0 tttO0 ett-0 ece0 8 ett-0 Cc £E 00S:0 00-0 OScO LILO tec-0 0Sc:0 l l 0001 000:1 l l HAGA! HAY COE Omeacee.0 See Omeer0 Il Ol l b p Lc-0 €800 LSt0 tce0 061:0 Ill-0 7c6c-0 1190 $c79O0 CSO E G E ete0 OSc0 070 tt£0 O00S:0 00:0 CLe0 OSc0 0Sc:0 l l 000:'| 000°! 0001 l l LOO SESIEO Se Z9T-0 SoC O Me coe0n meeee.0 CES Ol Febi0) mac ce:0 L910 491-0 L91-0 6 V6 6 uone[ndog 000-1 vrv-0 Z 1 I i saddjouas fo szaqunn u 8rl-0 WAS) = asx) gl d CSE0 00S:0" MEceE 0 “Sbe0 e) a 00S0 O0SO0 etttO ett0 V HdS Z (6 % i sadajouas fo 4aqunn 00S:0 OSc-0 a ccc'O =e 0 ©6008-0008 0 d ©) 8LC0 L990 OSc0 006-0 a V HCW i [ 1 if sadajouas fo saquny d e) 00S:0 0050 0001 000-1 ts 00S:0 00-0 V | WOd Z i i 1 sadajouad Jo saquny t60:0 L910 U SEE.O ec ee.0 al 940) (e330) d el) {3t2310) @) els) feheter(l) d Ding) sists) AS) fats) V | [Dd v t C I ey tts pur sno] SHIONANOAA ANAD - SNVLVN WAINOYNT 7 ATEVL 306 Q. O. N. KAY, R. F. JOHN AND R. A. JONES dispersal to occur between separate (although probably comparatively close) bodies of water. The results of our genctic analysis provide supporting evidence for the predominant réle of vegetative spread within lakes, canals and river systems, and also for the occurrence of seed-mediated dispersal between water bodics. Within lake and river systems our genetic analysis also provides evidence of the occurrence of reproduction by seed and consequent recombination in some areas, but the frequency of successful sexual reproduction and of allogamy cannot be determined because of the small number of variable loci and the probable occurrence of fixed heterozygosity, as described in the next section. GENETIC VARIATION A chromosome number of 2n = 42, probably hexaploid from a basic number of x = 7, has been reported for Swedish material of Luronium natans by Bjorkvist (1961), and a count of 2n = 38, which we cannot trace, was reported for British material by Clapham, Tutin & Warburg (1962), Suggesting that British material is alsc nexaploid. No counts on Bnitish or Insh material are documented in the Leicester Cytological Catalogue (R. Gornall, pers. comm. ). We obtained samples for isozyme studies of genetic variation from 16 populations or sub- populations in Wales, one of which was subdivided. A single sample was obtained from the newly-discovered population at Invermore Lough, Co. Galway (Connemara) in western Ireland. Details of the populations sampled are given in Table | and Figs. 1—4. Eleven loci in nine enzyme systems were assayed, of which only four were found to be variable, phosphoglucose isomerase 1 (PGI 1), phosphoglucomutase 1 (PGM 1), malate dehydrogenase (MDH) and shikimate dehydrogenase (SDH). Banding patterns were in agreement with the hexaploid chromosome counts reported for L. natans. Where all bands stained to the same intensity, it was assumcd that there were equal numbers of copies of each allele. Where there were differences in staining intensiies among the bands it was possible to estimate the number of copies of each allele that were present in an individual plant. Gene frequencies, together with the number of genotypes recorded at each site, are shown in Table 2. Two populations (Llyn Cwellyn and Invermore Lough) were represented by a single sample. Three of the 15 populations from which several samples were obtained (Llyn Eiddwen and Llyn Fanod, cach with six widely-spaced samples, and Ramsey West Pond, with ten samples from a dense pond population) were monomorphic at all variable isozyme loci. Two or more genotypes, with a maximum of four, were found in each of the twelve remaining populations. While each genotype could represent a true-breeding inbred linc, the strong vegetative spread and multiplication shown by the species suggest that apparent monomorphy, or paucity of genotypes within a population, is more likely to be the result of predominantly clonal reproduction. In a hexaploid, most isozymes will have three loci, so, in a survey of this type, it is not possible to distinguish true heterozygotes (individuals heterozygous at a single locus) from fixed heterozygotes (individuals which are homozygous for different alleles at corresponding loci in different genomes). Thus, although the observed frequency of heterozygosity was high, it was not possible either to confirm the occurrence of outcrossing and recombination, or to obtain an estimate of their frequency. The low ratios of numbers of genotypes found to numbers of variable loci (Table 2) and our examination of relative band intensities, which are normally proportional to the number of copics of each allcle that are expressed (c.g. Lack & Kay 1986) suggested that the frequencies of sexual reproduction, and especially of recruitment of new recombinant genotypes within a population, were low, and possibly zero in some cases. The distribution and frequencies of alleles at each scored locus are mapped in Figs. 1-4. The PG1 locus showed two major genotypes, one (found in Llyn Tegid, the Llangollen and Montgomery Canals, on Ramsey and in Ireland) with linked C and E alleles, and the other (from central and northern Walcs) with linked B and D alleles. The Irish sample from Invermore Lough was the only one which showed monomorphy and homozygosity at two of the loci, but many of the central and northern Welsh populations were monomorphic for probable fixed heterozygosity at two or more loci. RELATIONSHIPS AMONG POPULATIONS An unrooted phylogenetic tree of relationships among populations, based on the isozyme data and prepared using the CONTML procedure of Felsenstein’s PHYLIP package (Felsenstein 1993), is shown in Fig. 5, and Nei’s (1972) measures of genctic distance between populations, also based on the four variable loci, are shown in Table 3. The distances between populations along the branches LURONIUM NATANS IN BRITAIN AND IRELAND 307 ] 2 NA ny \ i La Jom 7 { \ ye NE a i, ape a= es 4 S (= Va ar ¢ NE ( Ya WZ a— ig AB, He a —s 12> N Ht a) Se St =e Se — QC Ne NT SN Bee 14 aT \ ? ss a — 13 — ESS ale — sf Be ES eee =e S ~ . fee a ee fi ae Nee A Seer Did Ss RG Lee y, ea ee 5 a ye ‘ ee S Ve N a Ireland <= A Ireland os WE ~ ff —_ ~ y UP a7 : Nie17 Ty \ ah =e : LZ =. i vail ~ 5S pis = 3 — ‘| Nee => Gare = i; A < 13 S77 11 ~~ AT — ~~ i aa = 16 ue) pe sa E> ee ee we Bi \ve 15 be iL—<—e> — NS } SS te i Le we, 7 . (iT a Ls Fe AA Spee, eee = ao ye et So = aS ae ree weer neg ae — ve . ~ Va (reiand wh BE ee pee —_ ~ ag a Ireland ee A ——_ o™ ~ 17 i ——ee— 17 Ficures 1-4. The distribution and frequency of alleles at four variable loci in 17 populations or sub-populations of Luronium natans in Wales and Ireland (inset). Sector size in each composite circular symbol is proportional to the frequency of the corresponding allele of each enzyme in the population sample from that site. PGI 1 is shown in Fig. | (top left), PGM 1 in Fig. 2 (top right), MDH in Fig. 3 (lower left) and SDH in Fig. 4 (lower right). Q. O. N. KAY, R. F. JOHN AND R. A. JONES 308 lpv-0 9S°:0 9S¢S:0 9tr-0 067-0 $860 859-0 OLV-0 ces 0 O0&L:0 81V-0 Olc:0 cee 0 L9v‘0 10¢:0 vel 0 69S°0 Li 0tL:0 OcI-1 tLv-0 688:0 €v0°1 0S8°0 1860 €c0°1 LS3:0 9tV-0 vis-0 cev0 119-0 1Sv-0 1$8-0 tvl:0 91 cLI‘0 901-0 9rV'0 601:0 ££0°0 £60:0 LII-0 Lt0-0 €81-0 vov'0 tve0 997:0 £9t-0 0¢0:0 €10°0 SI CLO ccv'0 vSc0 8L1°0 6$0:0 970:0 Lvl-0 9LV‘0 $6S ‘0 819-0 997-0 c9S ‘0 Iv1-0 S81-0 vl t6t 0 Lvc0 cv lO SSc-0 09¢:°0 vel 0 $90:0 1€é-0 S11-0 S1c-0 901-0 IS1-0 611-0 tl 9rS0 vsv-0 tlv-0 v9v'0 SLS‘0 8c 0 vvce-0 L6c:0 vOv'0 L9¢-0 ces ‘0 6SV-0 cl cS0:0 L91-0 691:0 $80°0 t6t-0 SeL-0 88:0 9S6:0 696°0 6L0:0 £60-0 Il ell-0 611-0 Sv0'0 857-0 vIs‘0 88:0 SLv-0 8Lt-0 ¢v0-0 610:0 01 Lc0:0 v80°0 60-0 e900 LtvO S$9¢0 ILvp-0 96S:0 £€LS-0 OcS:0 L8S0 vrr-0 SECON Cov: 0 -9LE-0 SOS0 ISO ccr0 €90:0 980:0 900:0 901:0 O€l:0 060-0 6 V6 8 Lel-0 vL0:0 1vc-0 Sel-0 6S$¢:0 961-0 LE C110 69:0 OL1:0 ces ‘0 Liv-0 9 (BIGUIDUUOZ) (|) Aaswiey 9] poue.4 UAT] ¢] uampprq UAT] P| € YloL uojy El V YlOL UOJ Tl UOIED S10D | | jueusyq UAT] OI @ YIOL UATT G6 V YIOL UAT] V6 JH UAT] 8 uOL.] ‘feuRD AJQWOSs]UO!|] /£ PpeoH S$, 8en ‘jeueg Alawo0s uo] 9 SOSSOID INO ‘;euRD AIOWOS]UO|| ¢ 691-0 jeued uajosuey] p 970-0 €bl-0 pisay Udy] ¢ 8Sv0 88£:0 09r:0 uayoremAq-A-uA]'] Z LI€EO 8S~-0 LEEO £400 uAyjam_ UATT | S v t c uo1eindog SNOLLV1NdOd DNOWYV FONVLSIC OLLANAD AO AUNSVANW (ZL61) S. IAN — SNVLVN WOINOUYNT € ATEVL LURONIUM NATANS IN BRITAIN AND IRELAND 309 Llyn Egnant Cors Caron Llyn Cwellyn ‘ Llyn Fanod Llyn Hir Montgomery Canal Four Crosses Llyn-y-Dywarchen \ \ 2 Afon Teifi .7 \ Llyn Tegid Liyn Teifi A - below ‘flash’ ¥ 4 ie Llangollen Llyn Teifi B Feeder Canal Mont Liyn Eiddwen a Saat Montgomery Canal Nag’s Head Ireland Invermore Lough Ramsey Island Afon Teifi above ‘flash’ Ficure 5. An unrooted phylogenetic tree, prepared using the CONTML procedure from PHYLIP (Felsenstein 1993), showing inter-relationships between 17 populations or sub-populations of Luronium natans. Distances between populations, measured along the arms of the diagram, are proportional to calculated genetic divergence. of the tree are proportional to calculated genetic divergence. While it must be remembered that this tree is based on only four variable loci, there is clear evidence of differentiation between geographic areas. The geographically isolated Ramsey and Connemara populations, which although well separated from one another are joined on a single branch of the tree, are also fairly widely separatedfrom the other populations, suggesting that they might be representatives (or relics) of one or more different western metapopulations. Ramsey Island, at the westernmost extremity of southern Wales, is actually closer to the Irish mainland (78 km) than to the nearest sampled populations of L. natans elsewhere in Wales (Llyn Fanod and Llyn Eiddwen, nearly 100 km to the north-west). Among the main group of Welsh populations, the populations from the Montgomery Canal and from Llyn Tegid (Bala Lake) and the Llangollen Feeder Canal form a relatively tight cluster in a central position on the tree, with the central and northern Welsh populations forming a much looser association. There is a surprisingly great genetic distance between the two subpopulations separated by the “Flash” on the Afon Teifi, perhaps as a result of downstream vegetative propagation of L. natans in the river current, from different older and effectively ancestral colonies upstream. The centre of genetic diversity of the species shown by the daia within Wales lies in the upland pools and lakes around Llyn Teifi, which drain into the Afon Teifi. Considering the pattern of genetic variability in more detail, there is some evidence of a cline of reducing genetic variability both northwards and westwards from the centre of distribution of the species in central Wales. The outermost populations have lower numbers of variable loci and lower numbers of genotypes than more central populations, although only a single individual could be found in the most northern population (Llyn Cwellyn) so there was no possibility of detecting variation in this population. However, each of the outlying populations in central and south-western 310 Q. O. N. KAY, R. F. JOHN AND R. A. JONES Wales (14—16) was represented by reasonable numbers of samples, and each population consisted of a single genotype, different in each case. At the PGI 1 and MDH loci, maximum variability was found in central Wales. Because of the probable predominance of clonal reproduction in this species, it was not possible to carry out statistical correlation tests to quantify genetic erosion, but as it is likely that clones can survive for very long periods of time, potentially deleterious effects of genetic erosion would be masked for correspondingly long periods. DISCUSSION Our genetic analysis shows that substantial genetic variation exists among Welsh populations of Luronium natans as a whole, but within-population variation was usually limited. The lowest within-population variation was found in the very isolated Ramsey population (ten samples) and in the comparatively isolated but extensive Llyn Fanod and Llyn Eiddwen populations (six samples in each case) which each consisted of only a single genotype, different at each site. The greatest genetic variation and within-population diversity were found in Llyn Teifi and its surrounding lakes, which may form a particularly effective metapopulation group (that 1s, a cluster of sites for a species, all more or less self-contained, but linked genetically over time). Nevertheless, even here the number of genotypes that we found was small, suggesting that clonal reproduction predominates. Several genotypes are also present in the “canal” populations from the much younger (less than 200 years) habitat of the Montgomery and Llangollen Canals. However, these populations, especially the Four Crosses sample, are genetically close to the probably large and ancient Llyn Tegid lake population. As Llyn Tegid (Bala Lake) is connected to the Montgomery Canal via the Afon Dyfrdwy (River Dee) and the Llangollen Feeder Canal, it seems extremely likely that the canal populations, and their genetic variation, have been derived from the Llyn Tegid population, in agreement with the suggestions of Lousley (1970) and Willby & Eaton (1993) that the canal populations of the species in the Welsh Borders and the lowland industrial areas of central and north-western England had originated in upland Wales. Canal traffic and water flow, and the capacity of the species for rapid vegetative extension and multiplication, would have facilitated their subsequent, documented spread along the full extent of the Montgomery and Shropshire Union canal (Briggs 1988), ultimately to more than 125 km from Llyn Tegid, and perhaps also even further afield into the canals of central and north-western England. The isozyme variability of Luronium natans contrasts with the extremely low or nil levels of isozyme polymorphism reported for other taxa in the same family (Alisma spp., Tnest 1991, Triest & Roelandt 1991, and Baldellia ranunculoides, Triest & Vuille 1991) and emphasizes the importance of case studies of individual species and populations in conservation genetics and conservation practice, rather than generalizations and assumptions based on apparently similar situations. In L. natans each population or distinct metapopulation 1s likely to have its own genetic characteristics, and should thus be regarded as a separate unit in any plans for conservation. The Welsh populations of Luronium natans (Figs 6 and 7) are probably one of its chief remaining strongholds within its world range, exhibiting the full spectrum of ecological diversity and, as we have shown, comparatively high levels of genetic variation. Their conservation and survival thus have particular importance. The declining or threatened status of L. natans across its world range, and the distribution, ecological status and importance of its native populations in Britain and Ireland, have not been fully appreciated in the past. In Britain, its colonization of abandoned canal systems led to the view that it was increasing (e.g. Rose 1983) and therefore had low conservation priority (Perring & Farrell 1983) although, in reality, expansion into canals probably did little more than to compensate for earlier losses from lowland pools and perhaps rivers. The observations of Briggs (1988), Willby & Eaton (1993) and Trueman et al. (1995) indicate that these canal populations are probably rather unstable and potentially transient. In addition to the deleterious efffects of competition, eutrophication and succession, Murphy & Eaton’s observations (1983) suggest that L. natans might in any case be eliminated from most of its existing canal sites if the canals were fully reopened for navigation by pleasure-boats. This presents something of a dilemma for the Countryside Council for Wales, English Nature and British Waterways, which are the agencies responsible for the conservation of Luronium natans, LURONIUM NATANS IN BRITAIN AND IRELAND @ 1970 onwards & 1930-1970 — Betore 1930 Ficure 6. The distribution of Luronium natans in 5 km squares of the National Grid in Wales and the Welsh Borders. See Table 1 for details of individual populations; several 5 km squares contain more than one population. Map prepared Ficure 7. The distribution of sampled populations of Luronium natans in 5 km squares of the National Grid in Wales. See Tables 1 and 3 for more details; some 5 km squares contained more than one sam- pled population. Map prepared using DMAP. using DMAP. but which also face strong pressure from recreational interests favouring re-opening or increased use of the canals for pleasure-boats. On the one hand, it can be argued that its spread into the canal habitat has reversed range decline for L. natans in lowland Bnitain, and that this habitat may now hold the largest population of the species anywhere in the world. On the other hand, this is a recent, highly artificial, and perhaps intrinsically transient distribution. British Waterways have attempted to conserve L. natans in a series of “off-line” canal reserves, and propose restoration of the Guilsfield Arm of the Montgomery Canal for the same purpose, but site management and population maintenance have presented considerable difficulties (Briggs 1996) and the long-term persistence of L. natans in such sites will probably depend largely on an appropriate interventionist management regime. The results of this study provide a new insight into this problem. They show that, despite the considerable abundance of Luronium natans in the Welsh canal system in which it occurs, there is comparatively little genetic difference between plants from different parts of the canal system. Indeed, they could probably all have been raised from a small proportion of the plants in Llyn Tegid and, given appropriate management, this could take place in a relatively short space of time. The whole series of Welsh canal populations could thus be seen as a subpopulation of the plants of one upland lake (Llyn Tegid). If correct, this interpretation indicates that the canal populations are of correspondingly less importance in terms of genetic conservation. The origins of the large population of L. natans centred on the canals of Manchester in north-western England, which we did not sample, are less clear, but we suspect that the same basic principles of low genetic diversity will apply, even if this population did not derive from the same source. In contrast, the long- established oligotrophic lake and river populations and metapopulations, and the surviving pond population on Ramsey, have been shown to have distinct genetic identities, markedly different from those in similar habitats elsewhere, and in a few cases even quite sharply different from closely adjacent sites. Within the main series of native populations in Wales, the highest priority should be given to the conservation of the native populations or metapopulations that show the greatest genetic diversity and of those that are long-established but disjunct. Examples of populations that grow in unusual S12 Q. O. N. KAY, R. F. JOHN AND R. A. JONES TABLE 4. LURONIUM NATANS —- CONSERVATION STATUS, DATES OF RECORDS AND HABITATS OF ALL KNOWN SITES IN WALES Vice-county and site Grid Status First Latest Habitat Status name reference record record ANGLESEY (52) Mynachdy Reservoir SH/31.92 1983 1987 Modified heath pond Probably extinct Llyn Dinam SH/31.77 SSSI 1813 1984 Lowland formerly Probably oligotrophic lake extinct Llyn Coron SH/37.70 SSSI 1813 1895 Lowland formerly Probably oligotrophic lake extinct Llyn Bodgylched SH/58.77 SSSI 1834 1834 Lowland formerly Probably oligotrophic lake extinct CAERNARFON (49) Llyn Glasfryn SH/40.42 SSSI 1987 1987 Lowland formerly Probably oligotrophic lake extinct Llyn Nantlle SH/Sileas 1834 1992 Upland oligotrophic lake —_Extant Llyn Cwellyn SH/56.54 SSSI 1895 1994 Upland oligotrophic lake —_ Extant Afon Rhythallt SH/54.63 1895 1967 Upland oligotrophic Unknown slow-moving river Llyn-y-Dywarchen SH/56.53 1895 1994 Upland oligotrophic lake —_ Extant Llyn-y-Gadair SH/56.52 1964 1992 Upland oligotrophic lake _—- Extant Llyn Padarn SH/57.61 SSSI 1848 1997 Upland formerly Extant oligotrophic lake Afon y Bala SH/585.601 LS 1985 Upland oligotrophic Probably slow-moving river extinct Llyn Peris SH/59.59 1805 1905 Upland oligotrophic lake —- Probably extinct Llyn Idwal SH/64.59 SSSI/ 1971 1971 Upland oligotrophic lake = Unknown NNR Llyn Cwmffynon SH/64.56 SSSI 1992 1992 Upland oligotrophic lake —_— Extant Llyn Llydaw SH/62.54 SSSI/ 197A 1971 Upland oligotrophic lake © Unknown NNR Afon Glaslyn SH/59.47 ? 1950 1950 Lowland oligotrophic Unknown slow-moving river MERIONETH (48) Llyn Cwmorthin SH/67.46 1961 1997 Upland oligotrophic lake —_— Extant Llyn Eiddew Bach SH/64.34 SSSI 1955 1955 Upland oligotrophic lake —_— Probably extinct Llyn Cwmbychan SH/64.31 SSSI 1921 1997 Upland oligotrophic lake —_— Extant Afon Eden SH/70.30 1960 1997 — Upland oligotrophic Extant slow-moving river Llyn Cynwch SH/73.20 1888 1996 Upland oligotrophic lake —_— Extant Llyn Tegid SH/89.31 SSSI 1805 1996 Upland oligotrophic lake —_ Extant DENBIGH (50) & FLINT (51) Llangollen Canal SJ/296.397 1862 1994 Lowland mesotrophic Extant —SJ/20.43 slow-moving canal 1. A ‘site’ is taken to mean a continuous, still or slow-moving water body; thus a canal population extending over 25 km or more is regarded as a single site, whilst two lakes linked by a fast-flowing stream would be listed separately. The names of the sites where one or more populations or subpopulations were sampled (see Table 3) are shown in italics. 2. ‘Lowland’ and ‘upland’ sites are defined by their catchment altitudes, so that Llyn Padarn at c. 100 m altitude is regarded as an ‘upland’ lake because of its primarily unenclosed catchment extending above 350m height, whilst Llyn Glasfryn at c. 130 m altitude remains wholly within a ‘lowland’ catchment. LURONIUM NATANS IN BRITAIN AND IRELAND 313 TABLE 4. CONTINUED Vice county and site Grid Status First Latest Habitat Status name reference record record MONTGOMERY (47) Montgomery Canal SO/12.94/- —_ SSSI 1933 1997 Lowland mesotrophic Extant $J/26.20 slow-moving canal Llyn Coch-hwyad SH/92.11 1993 1997 Upland oligotrophic lake —‘- Extant Llyn Gwyddior SH/93.07 1993 1997 Upland oligotrophic lake —_ Extant Llyn Bugeilyn SN/82.92 SSSI 1962 1995 Upland oligotrophic lake —- Extant Llyn Ebyr SN/97.88 SSSI 1988 1988 | Lowland mesotrophic Unknown lake CARDIGAN (46) Llyn-yr-Oerfa SN/72.79 1893 1893 Upland oligotrophic lake = Probably extinct Llyn Eiddwen SN/60.67 NNR 1893 1994 Upland oligotrophic lake —- Extant Llyn Fanod SN/60.64 SSSI 1893 1997 Upland oligotrophic lake —- Extant Llyn Teifi SN/78.67 SSSI 1893 1997 Upland oligotrophic lake —_ Extant Llyn Hir SN/78.67 SSSI 1989 1997 Upland oligotrophic lake —_ Extant Llyn Egnant SN/79.67 SSSI 1893 1996 Upland oligotrophic lake —_- Extant Llyn-y-Gorlan SN/78.66 SSSI 1893 1996 Upland oligotrophic lake _—‘ Extant Llyn Gynon SN/79.64 SSSI 1893 1994 Upland oligotrophic lake = Extant Afon Teifi SN/67.62 SSSI/ 1924 1997 —_ Upland mesotrophic Extant NNR slow-moving river RADNOR (43) Llyn Cerrig-Ilwydion SN/84.69 SSSI 1995 1997 Upland oligotrophic lake —_— Extant isaf Llyn Cerrig-Ilwydion SN/84.69 SSSI/ 1997 1997 Upland oligotrophic lake —‘- Extant uchaf NNR GLAMORGAN (41) Singleton SS/62.91 1840 1840 Lowland, site uncertain Probably extinct Crymlyn Bog SS/69.94 SSSI/ 1840 1840 Lowland oligotrophic fen Probably NNR extinct PEMBROKE (45) Porthmelgan stream SM/73.28 SSSI 1944 1944 Former heathland pond Probably extinct Dowrog Pool SM/77.27 SSSI 1905 1997 —_ Heathland pool Extant Penlan Farm reservoir SM/74.25 1981 1982 Modified heathland pond Unknown Ramsey Ponds SM/70.23 SSSI Is) 1997 Heathland pond Extant Houghton Farm SM/98.07 1982 1982 Lowland eutrophic pond Probably reservoir extinct or particularly threatened habitats (rivers and heathland ponds) should also be conserved where possible. The lake and pool habitats are delicately balanced and face a variety of threats, including eutrophication, acidification by acid rain and run-off from extensive conifer plantations, reservoir construction, pollution and disturbance, especially by powered boats and other recreational uses. At the 48 sites where it has been recorded in Wales (Table 4), L. natans is certainly or probably still - present (recorded in 1992 or later) in 28, of uncertain present status in six, and believed lost in 14 (five of these losses having taken place since 1980). By far the most severe decline has been in lowland sites. Ten out of 17 recorded populations (59%) have been lost here, as opposed to only four of the 31 upland populations (11%). The true ratio is likely to be worse, since in lowland sites Luronium natans is comparatively easy to detect, and there were certainly more lowland sites for the species in the past, which were not recorded 314 Q. O. N. KAY, R. F. JOHN AND R. A. JONES individually (e.g. Davies 1813), whereas recent searches in upland lakes, where L. natans is often hard to detect, have yielded a number of rediscoveries and also some new Sites. The selection of sites and species for conservation has tended to rely on records of range and abundance (measured, all too often, in terms of 10 km square distribution). This study shows not only how changeable these data can be, but also what other levels of significance are being overlooked. In order to conserve the species as a whole we need to consider more than just its overall numbers and where it occurs: we need to see the relationships within and between these data — a population (if not a metapopulation) level of analysis. If, in this sense, the canal sites for Luronium natans do not quite have their former priority, they still have considerable ecological significance. One consequence of this study is to show the relationship between an apparently remote and isolated upland locality for the species and a newly-created artificial lowland site. The rapid expansion into the canal system demonstrates the possibility of recovery for this species in the lowlands, and the importance of low-nutrient (now mainly upland) refuge sites. A strategy for the conservation and recovery of L. natans needs to integrate maintenance of refugia with the restoration of natural lowland habitat. It is to be hoped that future population management will take account of both processes, within a framework of applied genetic research. ACKNOWLEDGMENTS We are grateful to Tim Blackstock, Jonathan Briggs, Susan Byrne, Arthur Chater, David Evans, Stephen Evans, Geoffrey Halliday, Chris Preston, Tim Rich, Sue Scott, David Stevens, Max Wade, Sarah Whild and Nigel Willby for their generous advice and assistance, to Daniel Kelly for his careful and helpful editorial comments and corrections, and to the Countryside Council for Wales, which commissioned and supported the work described in this paper as part of a wider study of the genetics and conservation of scarce and declining plant species in Wales (Kay & John 1995). REFERENCES Baker, J. G. (1885). The Flora of the English Lake District. George Bell, London. Benoit, P. & RicHarps, M. (1963). A contribution to a Flora of Merioneth, 2nd ed. West Wales Naturalists’ Trust, Haverfordwest. ByorkKvistT, I. (1961). Luronium natans (L.) Raf. aterfunnen 1 Skane. Botaniska notiser 114: 365-367. Briccs, J. D., ed. (1988). Montgomery Canal ecological survey, Survey Report 1985-1988. Montgomery Canal Ecological Survey, Canal Wharf, Llanmynech. Briacs, J. (1996). Canals — Wildlife value and restoration issues. British wildlife 7: 365-377. Cuater, A. O. (1990). Recording in v.c. 46 during 1988 and 1989. Botanical Society of the British Isles, Welsh bulletin 49: 9-12. CrLaPpHaM, A.R., Tutin, T.G. & WarBurc, E.F. (1962). Flora of the British Isles, 2nd ed. Cambridge University Press, Cambridge. Davies, H. (1813). Welsh botanology. W. Marchant, London. Drisco.t, R. J. (1992). A further note on Floating Water-plantain, Luronium natans, in Norfolk. Transactions of the Norfolk and Norwich Naturalists’ Society 29: 184. Euis, R. G. (1983). Flowering plants of Wales. National Museum of Wales, Cardiff. FELSENSTEIN, J. (1993). PHYLIP — Phylogeny Inference Package (Version 3.5c). University of Washington, Seattle. Fercuson, C. M. (1991). The distribution of the Floating Water-plantain (Luronium natans (L.) Raf.) in Great Britain and Ireland. 40 pp. B.Sc. thesis, Loughborough Institute of Technology. Fercuson, C., Briccs, J. & WittBy, N. (1998). Floating Water-plantain in Britain — under-recorded and overlooked? British wildlife 9: 298-303. Fritz, O. (1989). Flytsvalting, Luronium natans, funnen i Halland 1988. Svensk botanisk tidskrift 83: 135-136. GrirFitH, J. E. (1895). The Flora of Anglesey and Carnarvonshire. Nixon & Jarvis, Bangor. Gutcu, J. W. G. (1842). A list of plants met with in the neighbourhood of Swansea, Glamorganshire. Phytologist 1: 180-187. Hauipay, G. (1978). Flowering plants and ferns of Cumbria. University of Lancaster, Lancaster. Hatuipay, G. (1997). A Flora of Cumbria. Centre for North-West Regional Studies, University of Lancaster, Lancaster. LURONIUM NATANS IN BRITAIN AND IRELAND a5 Hanspacn, D. & Krauscu, H. D. (1987). Zur Verbreitung und Okologie von Luronium natans (L.) Raf. in der DDR. Limnologica (Berlin) 18: 167-175. Hopcson, W. (1898). Flora of Cumberland. W. Meals, Carlisle. Hooker, W. J. (1831). The British Flora. Longman, Rees, Orme, Brown & Green, London. Hype, H. A. & Wane, A. E. (1957). Welsh flowering plants, 2nd ed. National Museum of Wales, Cardiff. Jones, R. A. & Ricn, T. C. G. (1998). Luronium natans / Baldellia ranunculoides / Alisma, in Ricu, T. C. G., Jermy, A. C. & Carey, J. L., eds. Plant crib 1998, pp. 314-317. Botanical Society of the British Isles, London. Kay, Q. O. N. & JouN, R. F. (1995). The conservation of scarce and declining plant species in lowland Wales: population genetics, demographic ecology and recommendations for future conservation in 32 species of lowland grassland and related habitats. Science Report no. 110, Countryside Council for Wales, Bangor. Lack, A. J. & Kay, Q. O. N. (1986). Phosphoglucose isomerase (EC 5.3.1.9) isozymes in diploid and tetraploid Polygala species: evidence for gene duplication and diversification. Heredity 56: 111-118. Locxton, A. & WILD, S. (1997). Rare plants of Shropshire: a Red Data Book of vascular plants. Shropshire Flora Group, Shrewsbury. Lous.ey, J. E. (1970). The influence of transport on a changing flora, in PERRING, F., ed. The Flora of a changing Britain (B.S.B.I Conference Report no. 11). E. W. Classey, Faringdon. Murphy, K. J. & Eaton, J. W. (1983). Effects of pleasure-boat traffic on macrophyte growth in canals. Journal of applied ecology 20: 713-729. Ne, M. (1972). Genetic distances between populations. American naturalist 106: 385-398. PERRING, F. H. & FARRELL, L. (1983). British Red Data books: 1. vascular plants, 2nd ed. Royal Society for Nature Conservation, Lincoln. PERRING, F. H. & Wa ters, S. M. (1976). Atlas of the British Flora, 2nd ed. EP Publishing, East Ardley. Preston, C. D. (1994). Luronium natans (L.) Raf., in STEWART, A., PEARMAN, D. A. & Preston, C. D., eds. Scarce plants in Britain, pp. 247-248. JNCC, Peterborough. Preston, C. D. & Crort, J. M. (1997). Aquatic plants in Britain and Ireland. Harley Books, Colchester. Preston, C. D. & Hitt, M. O. (1997). The geographical relationships of British and Irish vascular plants. Botanical journal of the Linnean Society 124: 1-120. Rico, T. C. G., Kay, G. M. & Kirscuner, J. (1995). Floating water-plantain Luronium natans (L.) Raf. (Alismataceae) present in Ireland. /rish Naturalists’ journal 25: 140-145. Ripey, H. N. (1930). The dispersal of plants throughout the world. L. Reeve, Ashford. RODRIGUEZ-OUBINA, J. & Ortiz, S. (1991). Luronium natans (Alismataceae) in the Iberian peninsula. Willdenowia 21: 77-80. Roe ors, J. G. M. (1983). Impact of acidification and eutrophication on macrophyte communities in soft waters in the Netherlands. I. Field investigations. Aquatic botany 17: 139-155. Rose, F. (1983). The wild flower key. Warne, London. SALTER, J. H. (1935). The flowering plants and ferns of Cardiganshire. University of Wales Press, Cardiff. SHaw, C. R. & Prasap, R. (1970). Starch gel electrophoresis of enzymes — a compilation of recipes. Biochemical genetics 4: 297-320. SHIELDS, C. R., OrTON, T. J., & STUBER, C. W. (1983). An outline of general resource needs and procedures for the electrophoretic separation of active enzymes from plant tissues, in TANKSLEY, S. D. & Orton, T. J., eds. lsozymes in plant genetics and breeding, Part A, pp. 443-468. Elsevier, Amsterdam. Situ, J. E. (1828). The English Flora, 2nd ed. Longman, Rees, Orme, Brown and Green, London. TriesT, L., ed. (1991). Isozymes in water plants. Opera botanica Belgica 4: 1-259. National Botanic Garden of Belgium, Meise. TrIEST, L. & ROELANDT, B. (1991). Isozymes in diploid and polyploid Alisma species (Alismataceac), in TRIEST, L., ed. Isozymes in water plants. Opera botanica Belgica 4, pp. 27-36. National Botanic Garden of Belgium, Meise. Triest, L. & VuILLeE, F.-L. (1991). Isozyme variation in several seed collections and hybrids of Baldellia (Alismataceae), in Triest, L., ed. Isozymes in water plants. Opera botanica Belgica 4, pp. 37-48. National Botanic Garden of Belgium, Meise. TRUEMAN, I., Morton, A. & WatnwriGut, M. (1995). The Flora of Montgomeryshire. Montgomeryshire Field Society and Montgomeryshire Wildlife Trust, Welshpool. TURNER, D. & Dittwyn, L. W. (1805). The botanist’s guide through England and Wales. Phillips and Fardon, London. Wiisy, N. J. & Eaton, J. W. (1993). The distribution, ecology and conservation of Luronium natans (L.) Raf. in Britain. Journal of aquatic plant management 31: 70-78. (Accepted January 1999) pores wou ~~ ~ ll a oa lo eel pees an ; 4 . , oe q ‘ i 4 oh ‘ Wade “a < ah te if 1 ihe i i 4 4) ha: il : ee =o Rea ht A Peek a i > \ q I ‘ip th ” ey Dad 4 : i : . - me — ~ Watsonia 22: 317-328 (1999) 317 Variation in the responses of infraspecific variants of wet grassland species to manipulated water levels BARRATT™*, D. R., WALKER, K. J., PYWELL, R. F.. MOUNTFORD, J. O. & SPARKS, T. H. Institute of Terrestrial Ecology, Monks Wood, Abbots Ripton, Huntingdon, Cambs., PE17 2LS ABSTRACT The effects of water levels on the germination, growth and reproductive effort of three wet grassland taxa were studied using an experimental water table facility. In addition, effects on related taxa including ecotypes and cultivars were examined. There were significant differences in biomass allocation and reproductive effort between infraspecific variants of Centaurea nigra and Lotus spp. in response to water table level. Variants had higher root and shoot dry-weights, produced more inflorescences and, in some instances, showed greater seedling establishment in the low water table treatments. In addition, significant differences in seedling establishment between infraspecific variants of both C. nigra and Lotus spp. were apparent. These findings have implications for the success of grassland re-creation schemes, and the selection of seed for such projects, and suggest that detailed research is needed to provide accurate information regarding the contrasting water table requirements of wetland plants, both at the species and subspecies levels. KEYWORDS: ecotypes, water regime requirements, wetland restoration. INTRODUCTION It is estimated that in the United Kingdom agriculturally improved grassland has increased by 90% in the past 50 years (H.M. Government 1995). Some of this increase has been at the expense of floristically diverse wet grassland communities which, in comparison, are of high conservation value. A possible technique for the re-creation of such habitats uses the sowing of seed to encourage the establishment of desirable plant species and communities (Wells, Cox & Frost 1989; Stockey & Hunt 1994). An objective of the present study was to assess whether commercially available seed is suitable for use in wet grassland re-creation projects. Purchased seed may vary in its provenance and in the accuracy with which species are identified. In many cases the provenance of commercial seed may mean that it is ill-suited to wetter sites: either as it has been taken from plants growing under much drier conditions (and may therefore exhibit ecotypic variation), or it has been specifically bred, in order to increase vigour and agronomic yield, as a cultivar. Material may be accurately named at the species level, but there may be uncertainty as to the precise subspecies involved. This imprecision may be important where two or more subspecies have significantly different responses to environmental variation. A similar problem may arise where the taxonomy of a group is difficult, e.g. where only specialists are able to distinguish microspecies within a complex. For example, seed supplied as Taraxacum officinale may include numerous microspecies, some of which differ in their water table requirements. Concern has also been expressed about the use of non-native genotypes of wildflower species in habitat restoration, and it has been postulated that genes from these non-native cultivars may dilute the gene pool of the native population (Cairns 1993; Akeroyd 1994). Non-native cultivars have often been specifically bred for rapid growth (Bullard & Crawford 1995), and may effectively out-compete and eliminate their native counterparts. If the restoration of wet grassland ecosystems is to be reliable and successful in the longer term, it is important to test the response of native and non-native genotypes to differing water regimes. The experiment described within the present paper set out to examine the effects of water levels on the germination, growth and reproductive effort of native material of three species. These effects were compared with the observed responses to the same treatments of infraspecific variants of these species including cultivars, “ecotypes” and subspecies. The experiment was conducted using *Address for correspondence: E-mail: Dbar@ITE.ac.uk 318 BARRATT, D. R., WALKER, K. J., PYWELL, R. F.,. MOUNTFORD, J. O. & SPARKS, T. H. an experimental water table facility at the Institute of Terrestrial Ecology (ITE) Monks Wood, where allied experiments on a range of native flood meadow species (Walker et al. 1997) have been conducted as part of a wetlands research programme. MATERIALS AND METHODS SELECTION OF SPECIES The selection of species was based on their occurrence within lowland wet grassland communities, and in particular their importance as constituents of seed mixtures prescribed for the restoration of such communities. Commercial seed of native provenance of Lotus corniculatus (Common Bird’s-foot-trefoil), Rhinanthus minor (Yellow-rattle) and Centaurea nigra (Common Knapweed) were sown. Their performance was compared with hand-collected seed from native populations, a forage variety (commonly used as a fodder crop) and seed of closely related species more frequently associated with wet conditions. The variants selected are listed below, with their provenance and abbreviated codes used to annotate text, tables and figures: 1. Lotus corniculatus “commercial” (Lcc) — commercially available seed of English provenance; 2. Lotus corniculatus “forage” (Lcf) — commercial forage variety, unknown provenance; 3. Lotus pedunculatus (Lp) — seed hand collected from populations in Dorset; 4. Rhinanthus minor “commercial” (Rmc) — commercial seed, possibly of ssp. minor, which 1s more associated with drier, often calcareous, soils; | 5. Rhinanthus minor ssp. stenophyllus (Rmw) — seed hand-collected from populations at Wicken Fen National Nature Reserve (NNR) in Cambridgeshire; 6. Centaurea nigra “commercial” (Cnc) — commercially available seed of English provenance; ~— Centaurea nigra ““Wicken” (Cnw) — seed hand collected from Wicken Fen National Nature Reserve. The seeds utilised in the study were less than a year old and placed in cold storage at 4°C prior to use. EXPERIMENTAL DESIGN The seeds were sown within circular plastic pots, 21 cm deep x 17 cm diameter, filled with approximately eight litres of soil-based compost. The soil had a silty-loam texture, broadly representative in physical characteristics of soils typically found on lowland alluvial grasslands, and was heat sterilised prior to use in order to reduce the number of weed species and pathogens present. Each pot contained 10 seeds and was labelled to identify the species and seed source. Four replicate pots of each taxon were then placed within large fibreglass tanks, situated outdoors on a gravel bed, in which water levels could be maintained at a constant level. Sixteen fibreglass tanks were used in the study. The four treatments (water table levels) were as follows: TO — tanks with the water table at the same height as the soil surface within the pots T50 — tanks with water table 50 mm below the soil surface within the pots T100 — tanks with the water table 100 mm below the soil surface within the pots T150 — tanks with the water table 150 mm below the soil surface within the pots The four water level treatments were replicated within four blocks. Therefore, 16 pots per seed source were placed at each water level. The placement of pots within tanks, and tanks within blocks, was fully randomised in order to eliminate bias. Evaporation losses and algal growth were reduced by placing white polypropylene granules on the surface of the water within the tanks. RESPONSES OF WET GRASSLAND SPECIES TO VARIED WATER LEVELS 319 MONITORING GERMINATION AND SEEDLING DEMOGRAPHY The numbers of seedlings in each pot were recorded every seven days from initial sowing in early December 1995 until the end of May 1996. Due to the large number of seedlings present (>5000) it was impractical to follow the fate of individual seedlings. Consequently successive cohorts of individuals were recorded within each pot. On four occasions germination could not be recorded when severe weather resulted in snow and ice covering the soil surface. BIOMASS ALLOCATION After approximately six months the seedlings were thinned to a single individual in each pot in order that plants could be grown on for destructive sampling without root competition. These individuals were then left to grow until October 1996, when they were harvested. After being oven dried at 80°C for 24 hours, both root and shoot dry-weights were obtained. It was not possible to obtain dry-weights for either of the R. minor variants as there were no surviving seedlings in October 1996. REPRODUCTIVE EFFORT At the final harvest the number of inflorescences present was also recorded. For the purposes of recording, the term “inflorescence” was taken to mean the capitulum in C. nigra and the cymose heads of the Lotus species. All species had seeded by the time of harvesting. Owing to seedling mortalities, it was not possible to gather inflorescence data for the R. minor variants. STATISTICAL ANALYSIS The preliminary statistical analysis, which was based on tank mean values, involved an examination of the effects of water table on the germination and survival of seeds of individual taxa (species and infraspecific variants). The effects of water table were broken down into constituent linear and quadratic components. It was considered important to make a preliminary analysis of how each taxon was responding to treatments, prior to “ecotypic’’ comparisons being made between infraspecific taxa of the same species. For each taxon, the cumulative gains (i.e. germination) and losses (i.e. mortality) of seedlings within each of the four treatments were calculated for four different time periods; 1) 49 days, ii) 98 days, 111) 154 days, and iv) 228 days after seed sowing. Over the four periods, a calculation was made of the mean number of seedling gains and losses per treatment for each taxon. Further analysis, using ANOVA, was carried out to examine whether individual taxa were exhibiting a significant response to treatment. The second stage in the statistical analysis involved an examination of whether closely allied taxa showed any marked differences in their response to the four water level treatments. Using data for the cumulative gains and losses of seedlings, (49, 98, 154 and 228 days after the seeds were sown), species were analysed simultaneously using split-plot ANOVA. Comparisons of the germination and survival of seedlings were made between the following taxa: Cnc and Cnw; Lcc, Lcf and Lp; and Rmc and Rmw. Further analysis of the variation in the responses of the infraspecific variants to water table depth was carried out using data on the establishment of seedlings, i.e. cumulative seedling germination minus cumulative seedling mortality. Possible differences in the establishment of taxa in the different treatments were examined using ANOVA. By October 1996, the time of the final harvest, many of the seedlings had died and consequently it was not appropriate to analyse the root, shoot and inflorescence data using ANOVA. An alternative, and more conservative method used regression of the four water table means, weighted by the number of valid observations contributing to those means. The results were shown as an estimate of the slope of the variable on water table (+ standard error) and an indication of significance. 320 BARRATT, D. R., WALKER, K. J., PYWELL, R. F.,. MOUNTFORD, J. O. & SPARKS, T. H. TABLE 1. SIGNIFICANCE OF THE LINEAR EFFECT OF WATER TABLE DEPTH ON THE GERMINATION, MORTALITY AND ESTABLISHMENT OF RMC, RMW, CNC, CNW, LCC, LCF AND LP SEEDLINGS OVER A 228 DAY PERIOD. Species Seedling Seedling Seedling Germination Mortality Establishment (Days after seeds were sown) (Days after seeds were sown) 49 98 154 228 49 98 154 228 228 Cnc ns ns - - ns ns ns ns - Cnw ns ns ns ns ns ns ns ns -- ck ns ns ns ns ns ns ns ns - ce ns ns ns ns ns ns + Lp ns ns (-) (-) ns ns ++ ++ so Rmc ns ns ++ + ns ns + +++ ns Rmw ns ns ns ns ns ns ns ++ ns Significantly higher rates of seedling germination, mortality and establishment in the highest water table levels are shown by: (+) = p<0-10; + = p<0-05; ++ = p<0-01; +++ = p<0-001. Significantly higher rates of seedling germination, mortality and establishment in the lowest water table levels are shown by: (-) = p<0-10; - = p<0-05; -- = p<0-01; --- = p<0-001. (ns = not significant). RESULTS EFFECTS OF WATER TABLE LEVELS UPON SEED GERMINATION, MORTALITY AND ESTABLISHMENT Table 1 shows the significance of the linear effects of water table depth on the germination, mortality and establishment of Rmc, Rmw, Cnc, Cnw, Lec, Lcf and Lp over a 228 day period. Rhinanthus minor Wicken (Rmw) and Rhinanthus minor commercial (Rmc) Seed of both taxa appeared to germinate more readily in the higher water table tanks. Germination was greatest 98-140 days after the seeds had been sown, after which time the rate of seedling mortality began to exceed germination, and hence net establishment of seedlings declined. After 228 days seedling establishment (Rmw and Rmc) had fallen to virtually zero in all four treatments. Cumulative germination of Rmc was found to be significantly larger in the higher water table treatments at two intervals: 154 days (p = 0-006) and 228 days (p = 0-010) after sowing. Cumulative germination of Rmw was not found to vary significantly between treatments. Cumulative seedling mortality after 228 days was, for both variants, significantly greater in the higher water table treatments. This offset the greater rate of germination in these tanks and consequently the overall establishment of Rmc and Rmw was not found to vary significantly between the different treatments. When comparing cumulative germination and cumulative seedling mortality, a marginally significant difference (p = 0-096) between the response of the two Rhinanthus taxa was noted 98 days after the seeds were sown. The germination of Rmc seed was found to be slightly greater than that of seed collected from Wicken Fen. This result may be explained by the negligible germination of Rmw seeds in the first fourteen weeks after sowing. In the final two recording periods, in which germination of both Rmc and Rmw increased, the difference in the cumulative germination of commercial and Wicken material was not found to be significant. Centaurea nigra commercial (Cnc) and Centaurea nigra Wicken (Cnw) In contrast to the two R. minor variants, the seeds of Cnc and Cnw were found to have higher cumulative germination levels in the lower water table treatments. This higher germination in the lower water table tanks was found to be significant for Cnc but not the Wicken variant. Cumulative germination and mortality of Cnc, in all four treatments, is shown in Figure 1. The final establishment of seedlings (i.e. cumulative germination minus cumulative mortality RESPONSES OF WET GRASSLAND SPECIES TO VARIED WATER LEVELS 321 —2— 0 mm gain —O— -50 mm gain — O— -100 mm gain - 1 ---150 mm gain —@=——() mm loss —®— -50 mm loss — @-— -100 mm loss -- @ ---150 mm loss Cumulative germination (gains and losses) -30 Days (after seeds were sown) FicurE |. Cumulative germination and mortality of Centaurea nigra commercial seedlings at four water table levels over a 228 day period. 0 mm = water table at soil surface level, -150 mm = water table 150 mm below soil surface. after 228 days) was significantly higher in the low water table treatments for both Cnc and Cnw (Fig. 2). However, a highly significant difference (p = < 0-001) was observed between the final establishment of Cnc and Cnw with a greater establishment of Wicken seedlings in three of the four treatments (Fig. 2). Lotus corniculatus commercial (Lcc), Lotus corniculatus forage (Lcf) and Lotus pedunculatus (Lp) Over the 228 day recording period, the establishment of Lcc and Lcf seedlings was erratic due to variable rates of seedling germination and mortality. Both seed types had higher germination and establishment rates in the lower water table treatments, but this was not significant at the p<0-05 level. Cumulative germination in all four treatments was considerably higher from seed of Lcc than from Lcf. The higher rate of Lcc germination was to some extent offset by seedling mortalities, which were higher than those experienced by Lcf. Net establishment of Lcf seedlings was, nevertheless, lower than that of Lcc in all four treatments. As found with Lcc and Lcf, there was a tendency for higher germination rates of L. pedunculatus in the lower water table treatments. This, combined with greater seedling mortality in TO (the high water table tanks), led to a significantly higher establishment of Lp in the low water table tanks (p = <0-001). Highly significant differences (p = <0-001) in the final establishment of Lcc, Lcf and Lcp were observed 228 days after the seeds were sown. The final establishment of Lp was considerably higher than that of Lcf in all treatments, and also greater than that of Lcc in treatments T100 and TSO CSS) EFFECTS OF WATER TABLE LEVELS UPON PLANT GROWTH AND REPRODUCTIVE EFFORT Tables 2 and 3 show the mean root and shoot dry weights, and mean number of inflorescences for variants of C. nigra and Lotus spp. at the four different water table levels studied. Centaurea nigra commercial (Cnc) and Centaurea nigra Wicken (Cnw) Mean root dry-weights of both C. nigra variants were significantly higher in the lower water table treatments. In the highest water table treatments, both variants had a mean root dry-weight of approximately 3 g, compared to mean root dry-weight in the lowest water table treatment of c. 56 g for Cnw and c. 64 g for Cnc. Cnw had slightly higher mean root dry-weights than Cnc in TO, T50 and T100. 522 BARRATT, D. R., WALKER, K. J., PYWELL, R. F., MOUNTFORD, J. O. & SPARKS, T. H. Centaurea nigra A 60 50 40 30 20 Seedling establishment 10 0 50 100 150 200 250 Number of days (after seeds were sown) Centaurea nigra B —— mm —eo— -50 mm — #— -100 mm -- @---150 mm Seedling establishment 0 50 100 150 200 250 Number of days (after seeds were sown) FiGure 2. Comparison of the establishment (germination minus mortality) of seedlings of Centaurea nigra commercial (A) and Centaurea nigra Wicken (B) at four water table levels over a 228 day period. RESPONSES OF WET GRASSLAND SPECIES TO VARIED WATER LEVELS 323 Seedling establishment 0 50 100 150 200 250 Seedling establishment 0 50 100 150 200 250 Seedling establishment 0 50 100 150 200 250 Days (after seeds were sown) Ficure 3. Comparison of the establishment (germination minus mortality) of seedlings of Lotus corniculatus forage (A), Lotus corniculatus commercial (B) and Lotus pedunculatus (C) at four water table levels over a 228 day period (note variations in scale). 324 BARRATT, D. R., WALKER, K. J., PYWELL, R. F., MOUNTFORD, J. O. & SPARKS, T. H. TABLE 2. MEAN ROOT AND SHOOT DRY WEIGHTS, AND MEAN NUMBER OF INFLORESCENCES OF CENTAUREA NIGRA WICKEN (Cnw) AND CENTAUREA NIGRA COMMERCIAL (Cnc) SEEDLINGS AT FOUR DIFFERENT WATER LEVEL TREATMENTS (0, 50, 100 AND 150 MM BELOW THE SOIL SURFACE). Mean root dry weights (g) Mean shoot dry weights (g) Mean no. of inflorescences Treatment Cnc Cnw Cnc Cnw Cnc Cnw TO 3-07 3-58 2-00 2-62 0-00 0-00 T50 4.98 7:21 3-53 3-80 0-33 0-37 T100 31-08 36-15 11-51 10-54 3-94 3-20 Tels 64-29 56-48 15-49 13-54 6-71 5-00 TABLE 3. MEAN ROOT AND SHOOT DRY WEIGHTS, AND MEAN NUMBER OF INFLORESCENCES OF LOTUS CORNICULATUS FORAGE (Lcf), LOTUS CORNICULATUS COMMERCIAL (Lec) AND LOTUS PEDUNCULATUS (Lp) AT FOUR DIFFERENT WATER LEVELS (0, 50, 100 AND 150 MM BELOW THE SOIL SURFACE). Mean root dry weights (g) Mean shoot dry weights (g) | Mean no. of inflorescences Treatment eck Lec Lp ch Ieee Lp ech ee Lp TO 0-00 1-19 0-65 0-00 1-83 0-70 0-00 14-00 0-00 TSO 12-56 1-01 2-47 5-43 1:50 4.44 114-00 37-25 18-33 T100 26:76 7-42 14-87 9-32 6-77 12-85 76-33 63-50 29-44 TSO 46-40 28-95 21-77 21-00 21-14 18:16 333-86 142-08 56-19 A similar trend was observed for mean shoot dry-weights, with marginally higher dry-weights from Wicken material in TO and TSO, and higher dry-weights of Cnc in T100 and T150. Again, both C. nigra variants had significantly higher shoot dry-weights in the lower water table treatments (Cnc p = 0-028, Cnw p = 0-034). Neither Cnc nor Cnw flowered in the highest water table tanks. The mean number of inflorescences, of both variants, was significantly higher in the lowest water table tanks (Cnc p = 0-029, Cnw p = 0-032). Lotus corniculatus commercial, Lotus corniculatus forage and Lotus pedunculatus Mean root dry-weights of the three Lotus taxa were significantly higher in the lower water table treatments. The highest mean root dry-weight recorded was 46-40g for Lcf forage in T150. Mean shoot dry weights also increased significantly as water table level decreased, with this trend being particularly notable for Lp (p = 0-008). Like the C. nigra variants, Lcc and Lp plants produced significantly more inflorescences in the lower water table treatments. Although not found to be significant, the number of Lcf inflorescences was also markedly higher in the lowest water table tanks than in the other three treatments. Indeed, in T150 Lcef produced substantially more inflorescences than the other two Lotus strains. Only Lcc flowered in TO, the highest water table treatment. DISCUSSION The effects of water stress upon plant physiology and germination have been well documented (e.g. Evans & Etherington 1990; Jackson 1990; Crawford 1996; Olsson et al.1996). A rise in water levels can deprive plants of oxygen, may affect the production and transport of plant hormones and can increase the likelihood of microbial attack (Crawford 1996). Some species have developed adaptations, both morphological and physiological, to help them cope with water table changes (e.g. Voesenek, Blom & Pouwels 1989). RESPONSES OF WET GRASSLAND SPECIES TO VARIED WATER LEVELS 325 If a plant becomes adapted to suit the specific environmental conditions of its habitat, such that its requirements become markedly different from those of other populations within the same species, ecotypic differentiation has occurred. The ability to select plants that are suited to environmental stresses, such as flooding, drought or even soil contamination, has led to the creation of new plant and crop varieties through plant breeding (e.g. Elias & Chadwick 1979; Yaseen & Al-Omary 1994) and also has large potential for providing guidance when selecting natural species for habitat restoration schemes. For instance, as L. corniculatus is a highly variable species (Jones & Turkington 1986) it would seem appropriate to sow the seed of a wetland “‘ecotype” in sites with higher water table regimes. RHINANTHUS SPP. Ter Borg (1985) studied the population biology of a number of hemi-parasitic Scrophulariaceae, and noted that Rhinanthus species often exhibit a wide infraspecific variation. R. angustifolius, for example, is thought to have at least eight different subspecies (Oberdorfer 1979). Taxonomic difficulties in Rhinanthus have contributed to the concept of seasonal dimorphism whereby species such as R. minor have been sub-divided into two taxa; aestival and autumnal, based upon the time of flowering and certain morphological characteristics. In some species of Rhinanthus, So6 (1970) has identified not only autumnal and aestival variants, but also montane, alpine and segetal variants (Karlsson 1974). Seasonal dimorphism, as a means of explaining variation in Rhinantheae, is now seen as an oversimplification with some of the apparent differences having been shown to have a genetic basis (Ter Borg 1985). Ecotypic variation may also have been overlooked when attempting to explain the apparent variations in the genus (Karlsson 1974). Grime, Hodgson & Hunt (1988) quote R. minor var. stenophyllus and R. minor var. minor as the two commonest R. minor ecotypes. R. minor var. minor is thought to prefer drier sites in southern England with R. minor var. stenophyllus being more suited to moist grasslands, particularly in the north. In this study, R. minor appeared to germinate more readily in the higher water table tanks, though this trend was only statistically significant for the Rmc (after 154 and 228 days) and not Rmw. Owing to high rates of seedling mortality, no significant differences in the final establishment of Rmec or Rmw could be demonstrated. It is recommended that, for future research, sufficient numbers of plants are established to ensure full representation at each water table depth. CENTAUREA NIGRA The findings for C. nigra were both fuller and more interesting. Seed of both Centaurea variants had higher rates of germination in the lower water table treatments. However, unlike Cnw, the germination of commercial C. nigra seed was significantly higher in the low water table tanks, suggesting that germination of Wicken seed may be less dependent upon water table than seed which is commercially available. This supposition is supported by the seedling establishment data, which showed the final establishment of Wicken seedlings to be significantly greater than the establishment of commercial seedlings (p<0-001). These results imply that Cnw seeds are not only more suited to wetter sites than commercial seed, but also to drier sites as well. The germination of seed originating from Wicken Fen was not significantly affected across the range of water tables studied. Therefore, seed collected from Wicken Fen and similar sites may, in terms of germination and establishment, be more suited for use in wet grassland and other restoration schemes than seed which is commercially available at present. Both C. nigra variants had significantly higher root and shoot dry-weights and number of inflorescences in the low water table tanks. In the T150 tanks, the commercial plants had slightly higher dry-weights, and produced more inflorescences, than the Wicken plants. C. nigra is, according to Grime et al. (1988), a complex group in need of further taxonomic and ecological study. Although botanists have often identified the subspecies nigra and nemoralis, the distinction between the two is no longer believed to be consistent (Stace 1991). It is likely that the wide distribution of C. nigra is, in part, linked to its considerable genetic variation (Grime et al. 1988). : LOTUS SPP. Seed of the two L. corniculatus variants showed lower germination rates in the high water table ianks with significantly higher overall establishment of seedlings in the low water table treatments. These results contrast with the findings of Baker (1988), where waterlogging had no effect on the 326 BARRATT, D. R., WALKER, K. J., PYWELL, R. F.. MOUNTFORD, J. O. & SPARKS, T. H. germination of either L. corniculatus or L. pedunculatus. In the present study, the effect of water table on the establishment of L. pedunculatus seedlings was found to be highly significant (p<0-001), with highest establishment in the low water table treatments. This observation makes interesting comparison with the usually observed habitat preference for Lotus spp. where L. corniculatus is thought to be extremely tolerant of water deficit (Grime et al. 1988; Bullard & Crawford 1996) whilst L. pedunculatus tends to grow in damp areas with a higher water table. Nevertheless, the findings do support the hypothesis of Blumenthal, Aston & Pearson (1996) that “variation exists within Lotus species for ability to germinate over a range of moisture potentials”, implying that the establishment of Lotus spp. in wet grassland restoration schemes may be markedly influenced by water table levels. Highly significant differences between the establishment of the three Lotus taxa were revealed. The establishment of Lcf was lower than that of the Lcc and Lp in all four treatments. In the two lowest water table treatments, L. pedunculatus also had markedly higher seedling establishment than L. corniculatus commercial. These results suggest that, in terms of seedling establishment. seed of Lcf would be less reliable for restoration purposes than Lcc and Lp seed. Sites with relatively low water tables might also benefit if L. pedunculatus seed were sown, as it was found to establish better than the commercial seed in treatments T100 and T150. The biomass data underlined the germination and seedling establishment findings, with all three Lotus species performing significantly better (with higher dry-weights) at the lower water table levels. Unsurprisingly, the mean root and shoot dry-weights of the forage variety were markedly higher than those of Lcc and Lp. When examining inflorescence data at the lowest water table treatment, Lcf produced almost six times as many inflorescences as Lp and twice as many inflorescences as Lec. This highlights the need to consider a range of factors, and not just seedling germination and establishment, when making decisions on the suitability of seed for use within restoration mixtures. Introducing forage and commercial cultivars, which may grow quickly and produce many inflorescences, could put less vigorous native species at a competitive disadvantage. Grant & Small (1996) have carried out a detailed examination of the ancestry of the L. corniculatus complex in which evidence regarding the geography, genetics, chemistry and morphology of the genus was reviewed. On the basis of this they have suggested that L. corniculatus may have arisen from a hybrid of L. glaber and L. pedunculatus. If L. pedunculatus is a direct parent of L. corniculatus this might partly account for the ability of L. corniculatus to germinate and establish in both damp and dry conditions. Following an examination of the response of 91 species to flooding, Justin & Armstrong (1987) classtfied L. corniculatus as a plant of intermediate habitats. Prior to this research, they had assumed it to be a species indicative of non-wetland environments. Such findings further underline the potential variability within this genus, which makes defining the precise water table requirements of any available strain of L. corniculatus difficult for restoration purposes. As the present study was based upon seed obtained from a small number of sources, only a small fraction of the genetic variation present within Lotus may have been examined. The present study has shown that the germination and establishment of seeds and seedlings, both between and within species, can vary significantly depending on the source of the seed. Seed collected from “ecotypes’, adapted to specific environmental conditions, may have significantly different hydrological requirements to those of seed collected from other populations. A recognition and understanding of the differing autecology of ecotypes, and other closely related species and varieties, is essential if wetland restoration schemes are to achieve their full potential. Seed obtained from commercial sources may, as demonstrated in the present study, perform poorly in comparison to seed obtained from other known populations. In addition, one is often uncertain as to the age of commercially supplied seed and the conditions in which it has been stored. A lack of accurate information regarding the identification (usually at the subspecies level) and precise origin of commercial seed is a further problem. Further research is required to investigate the effects of different water table regimes on both intra- and inter-specific competition between a range of ecotypes and cultivars of common wild flower species. Such interactions may exert considerable influence on the outcome of secondary succession following the sowing of seed mixtures containing such ecotypes and cultivars. It is likely that competitive interactions will be significantly accentuated under higher water table conditions. Using an experimental: water table facility, the effects of water levels on the RESPONSES OF WET GRASSLAND SPECIES TO VARIED WATER LEVELS 327 germination and survival of 23 wet grassland species have been studied (Walker et al. 1997). The first species to germinate were productive grasses, which suppressed the establishment of a number of herb species. This suggests that attempts to re-create wet grassland using seed may not always lead to the establishment of those species which germinate late and are subsequently out-competed. In such circumstances, plug plants may be a more effective means of introducing desirable species (Walker et al. 1997). Crawford (1996) has reiterated the need for a holistic approach when studying how plants react to water table fluctuations. Plants may vary in their response to water stress at different stages in their life cycle (Evans & Etherington, 1990; Yaseen & Al-Omary 1994; Ollson er al., 1996) and have to adapt to changing water levels whilst being in competition with other species (Walker et al. 1997). The present study was conducted under stable, non-competitive conditions and, as noted by White (1985), “there is doubt about the usefulness of studying the population dynamics of natural plant populations in isolation from their phytosociological environment”. CONCLUSIONS It is estimated that between 1930 and 1984 semi-natural lowland grassland decreased by 97% in England and Wales (Fuller 1987; H.M. Government 1995). Re-creation of such communities and habitats needs to be underpinned by a scientific understanding of wetland species and ecosystems. This preliminary experiment has demonstrated that the autecological requirements of closely allied taxa and infraspecific variants should not be overlooked. The biomass and reproductive effort of variants of Centaurea nigra and Lotus spp. were found to be significantly affected by water table depth. Variants had higher root and shoot dry-weights, produced more inflorescences and, in some instances, showed greater seedling establishment in the low water table treatments. In addition, significant differences in seedling establishment between infraspecific variants of both C. nigra and Lotus spp were apparent. It is probable that such differences also exist between “ecotypes” of other species and research is needed to further our understanding of the water table requirements of the main constituent species and subspecies of wet grassland vegetation (Mountford & Chapman 1993). Habitat re-creation will become more successful if species’ requirements are matched to site conditions. At present, the suitability of commercially available seed for wet grassland re-creation schemes is largely unknown and requires further investigation. ACKNOWLEDGMENTS We would like to thank the Ministry of Agriculture, Fisheries and Food (M.A.F.F.) for providing the funding to carry out this work. We are grateful to the National Trust who provided permission for seed to be collected from Wicken Fen National Nature Reserve, Cambridgeshire. Liz Warman (I.T.E., Monks Wood) and Caroline Ashton (formerly I.T.E., Monks Wood, now Farming and Rural Conservation Agency) provided valuable assistance in the collection of seed and the weekly recording of germination within the tanks. REFERENCES AKEROYD, J. (1994). Seeds of destruction? Non-native wildflower seed and British floral biodiversity. Plantlife, London. Baker, K. (1988). Ecospeciation in the genus Lotus. M.Sc. thesis, University of Newcastle. BLUMENTHAL, M. J., Aston C. & Pearson, C. J. (1996). Effect of temperature and moisture potential on germination and emergence in Lotus sp. Australian journal of agricultural research 47: 1119-1130. BULLARD, M. J. & CRAWFORD, T. J. (1995). Productivity of Lotus corniculatus L. (bird’s-foot trefoil) in the UK when grown under low-input conditions as spaced plants, monoculture swards or mixed swards. Grass and forage science 50: 439-446. BuLLarb, M. J. & CRAwrorD, T. J. (1996). Seed yield, size and indeterminancy in diverse accessions of Lotus corniculatus L. grown in the UK. Plant varieties and seeds 9: 21-28. Cairns, J. (1993). Disturbed ecosystems as opportunities for research in restoration ecology, in JoRDAN, W. R., Gitpin, M. E. & Asser, J. D., eds., Restoration ecology: 301-320. Cambridge University Press, Cambridge. 328 BARRATT, D. R., WALKER. K. J., PYWELL, R. F.. MOUNTFORD, J. O. & SPARKS, T. H. CrawrorD, R. M. M. (1996). Whole plant adaptations to fluctuating water tables. Folia geobotanica et phytotaxonomica 31: 7-24. Exias, C. O. & Cxapwick, M. J. (1979). Growth characteristics of grass and legume cultivars and their potential for land reclamation. Journal of applied ecology 16: 537-544. Evans, C. E. & ErHerincton, J. R. (1990). The effect of soil water potential on seed germination of some British plants. New phytologist 115: 539-548. Futter, R. M. (1987). The changing extent and conservation interest of lowland grasslands in England and Wales: A Review of Grassland Surveys 1930-1984. Biological conservation 40: 281-300. GranT, W. F. & Sma t, E. (1996). The origin of the Lotus corniculatus (Fabaceae) complex: a synthesis of diverse evidence. Canadian journal of botany 74: 975-989. Grime, J. P., Hopcson, J. G., & Hunt, R. (1988). Comparative plant ecology. A functional approach to common British species. Unwin Hyman, London. H.M. GoveRNMENT (1995). Biodiversity. The UK Steering Group Report. Volume 2: Action Plans. Cmd 2428. H.M.S.O., London. Jackson, M. B. (1990). Hormones and developmental changes in plants subjected to submergence or soil waterlogging. Aquatic botany 38: 49-72. Jones, D. A. & TuRKINGTON, R. (1986). Biological Flora of the British Isles No. 163. Lotus corniculatus L. Journal of ecology 74: 1185-1212. Justin, S. H. F. W. & Armstrona, W. (1987). The anatomical characteristics of roots and plant response to soil flooding. New phytologist 106: 465-495. Karisson, T. (1974). Recurrent ecotypic variation in Rhinantheae and Gentianaceae in relation to hemiparasitism and mycotrophy. Botaniska notiser 127: 527-539. MOounrTForD, J. O. & CHAPMAN, J. M. (1993). Water regime requirements of British wetland vegetation: Using the moisture classifications of Ellenberg and Londo. Journal of environmental management 38: 275-288. OsBeRDORFER, E. (1979). Pflanzensoziologische exkursionsflora. Ulmer, Stuttgart. Ovsson, M., Nicsson, K., LILJENBERG, A. & HENDRY, G. A. F. (1996). Drought stress in seedlings: Lipid metabolism and lipid peroxidation during recovery from drought in Lotus corniculatus and Cerastium fontanum. Physiologia plantarum 96: 577-584. Soo, R. (1970). Arten and Unterarten der Gattung Rhinanthus in Europa. Acta botanica Academiae Scientiarium Hungaricae 16: 193-206. Stace, C. A. (1991). New Flora of the British Isles. Cambridge University Press, Cambridge. StockEy, A. & Hunt, R. (1994) Predicting secondary succession in wetland mesocosms on the basis of autecological information on seeds and seedlings. Journal of applied ecology 31: 543-559. TeR Bore, S. J. (1985). Population biology and habitat relations of some hemiparasitic Scrophulariaceae, in Wulte. J., ed., The population structure of vegetation: Pp. 441-463. Dordrecht. VoESENEK, L. A. C. J., Blom, C. W. P. M. & Pouwe.s, R. H. W. (1989). Root and shoot development of Rumex species under waterlogged conditions. Canadian journal of botany 67: 1865-1869. Wacker, K. J., et al. (1997). Experimental investigations into the effects of water-levels on the germination and survivial of 23 wet grassland species, in Mountrorp, J. O. M., ed.. M.A.F.F. wetlands projects: Publication and dissemination of findings. Institute of Terrestrial Ecology (N.E.R.C.) Final Report to the Ministry of Agriculture, Fisheries and Food. WELLS, T. C. E., Cox, R. & Frost, A. (1989). Focus on nature conservation. No. 21 The establishment and management of wildflower meadows. Nature Conservancy Council, Peterborough. Wulte, J. (1985), ed., The population structure of vegetation. Dr W. Junk Publishers, Dordrecht. YASEEN, B. T. & AL-Omary, S. S. (1994). An analysis of the effects of water stress on leaf growth and yield of three barley cultivars. /rrigation science 14: 157-162. (Accepted January 1999) Watsonia 22: 329-342 (1999) 329 The distribution and habitat of Potamogeton X suecicus K. Richt. (P. filiformis Pers. x P. pectinatus L.) in the British Isles C. D. PRESTON ITE Monks Wood, Abbots Ripton, Huntingdon, Cambs., PE17 2LS P. M. HOLLINGSWORTH Royal Botanic Garden, Inverleith Row, Edinburgh, EH3 5LR and R. J. GORNALL Department of Biology, University of Leicester, Leicester, LEI 7RH ABSTRACT Recent studies of Potamogeton X suecicus K. Richt. (P. filiformis x P. pectinatus, Potamogetonaceae) have shown that the hybrid can be identified by morphological criteria and by isozyme analysis. The latter suggests that all plants studied are F, hybrids and does not provide any evidence for backcrossing to the parents. The hybrid is widespread in Britain in the area where both parents occur and also occurs in two rivers south of the current distribution of P. filiformis. In recent years it has been discovered at widely scattered localities in Ireland. The localities from which P. x suecicus has been recorded are listed, with records confirmed by isozyme analysis distinguished. In some rivers, streams and lakes it is the dominant macrophyte, growing in large, almost pure stands, whereas in other localities it is found as scattered plants in more open communities. The vegetation at the British sites is summarised from the results of 40 quadrats recorded at 21 sites. Sites where the hybrid may grow in the absence of one or even both parents are identified; they include rivers and shallow coastal lakes over sand. The history of the hybrid in Britain suggests that it is very easily over-looked; almost all records have been made in two periods (1940-1950, 1986-1998) when observers familiar with the plants were active on fieldwork. It is almost certainly still under-recorded in Scotland and Ireland. Keyworps: Aquatic vegetation, hybridisation, isozyme analysis, orphaned hybrids. INTRODUCTION The detailed studies of the genus Potamogeton which J. E. Dandy and G. Taylor carried out between 1936 and 1976 did much to clarify the taxonomy and distribution of the British and Irish species and their hybrids. Indeed, it might be argued that Dandy & Taylor accomplished almost as much as one could hope to achieve using the traditional methods of herbarium taxonomy, leaving to their successors only the publication of a detailed account of the British and Irish taxa (which Dandy & Taylor never completed) and the tying up of the loose ends which are inevitably left at the end of any botanical career. However, the ever-increasing range of molecular techniques which have become available to plant taxonomists since the 1960s have allowed many aspects of the subject to advance beyond the restrictions hitherto imposed by traditional methods. This paper arises from a study of the distribution and ecology of the widespread and taxonomically difficult hybrid Potamogeton xX suecicus which has combined traditional field and herbarium studies with the now well-established techniques of isozyme analysis. Potamogeton x suecicus K. Richt. is the hybrid between the two species of Potamogeton 330 C. D. PRESTON, P. M. HOLLINGSWORTH AND R. J. GORNALL Ficure 1. The distribution of Potamogeton X suecicus in the British Isles. Squares indicate 10-km grid squares where the hybrid has been seen from 1990 onwards; circles indicate squares where it was last seen before 1990. Subgenus Coleogeton which occur in the Bnitish Isles, P. filiformis Pers. and P. pectinatus L. In gencral appearance it is closer to P. pectinatus than to P. filiformis but unlike P. pectinatus at least some of the leaf sheaths are tubular at the base. The stigmas may be borne on a distinct style (as in P. pectinatus) or be sessile (as in P. filiformis). Plants of P. x suecicus may flower freely from May to October but their pollen is sterile and they do not set fruit. Certain identification of the hybrid is not possible in the field, but under the binocular microscope the hybrid can be identified with some confidence using morphological characters alone. All three taxa are described and illustrated in Preston (1995) and an updated key, modified to take account of the presence of the similar hybrid P. x bottnicus Hagstr. in Britain, is provided by Preston et al. (1998). The presence of Potamogeton X suecicus in the British Isles was first established by Dandy & Taylor (1940). They showed that earlier records of P. x suecicus in Britain were erroneous, but they DISTRIBUTION OF POTAMOGETON x SUECICUS 331 recognised the hybrid from a few herbarium specimens collected in Scotland in the 19th and early 20th centuries and from collections made in the Outer Hebrides by J. W. Campbell, W. A. Clark and A. J. Wilmott in 1938 and 1939. Further details of the Hebridean records, and additional records from the Hebrides and elsewhere, were published by Dandy & Taylor (1941, 1946), Clark (1943), Clark & Heslop Harrison (1940), Heslop Harnson (1941, 1949), Heslop Harrison & Clark (1941, 1942), Heslop Harrison & Heslop Harrison (1950) and Heslop Harmison et al. (1941, 1942). By 1950 P. Xx suecicus had been recorded in eight vice-counties, two (v.cc. 93 and 102) on the basis of old herbarium specimens and the rest (v.cc. 64, 65, 68, 81, 103 and 110) solely or in part on the basis of plants collected since 1940. Dandy & Taylor’s (1946) records from the River Wharfe and River Ure in Yorkshire were supported by the anatomical studies of Bance (1946) and were especially significant. These rivers lie south of the known distribution of the rarer parent, P. filiformis, although subfossil fruits of P. filiformis have been found in Late Glacial deposits in southern England (Godwin 1975) and the species persisted in Anglesey until 1826 (Preston 1990). Although many records of P. x suecicus were published between 1940 and 1950, there were very few additional discoveries during the next 35 years. The sites in the English rivers were often visited, and N. T. H. Holmes extended the distribution of the plant from the River Tweed into its tributary, the River Till (Holmes & Whitton 1975a, b). In Scotland M. McCallum Webster (1978) discovered the hybrid in Moray (v.c. 95) and U. K. Duncan (1969) rediscovered it at one of Heslop Harnison’s localities on Tiree (v.c. 103). The hybrid was mapped by Perring & Sell (1968) and covered by Dandy’s (1975) account of British and Irish Potamogeton hybrids. Since 1985 there has been renewed interest in P. x suecicus. In 1986 C.D.P. began preparatory fieldwork for the B.S.B.I. handbook Pondweeds of Great Britain and Ireland (Preston 1995), and with N. F. Stewart refound P. x suwecicus at several of the sites from which it had previously been recorded. In 1991 P.M.H. & R.J.G. began a detailed study of P. x suecicus and its parents, using isozyme analysis, in collaboration with C.D.P. (Hollingsworth et al. 1996a, b). As a direct or indirect result of this activity the hybrid has been rediscovered at most of its old localities, and several new sites discovered. These have included the first records from three vice-counties in Scotland, and the first records from Ireland. This paper draws together the recent records of the hybrid, documents the populations which have been identified on the basis of isozyme as well as morphological evidence, and describes the habitat of the hybrid in the British Isles. Although recent work has confirmed the identity of many populations of P. x suecicus, it has also indicated that the material from the River Till and River Tweed differs in both morphological and isozyme characters from the other plants studied, and is apparently referable to the hybrid between P. pectinatus and P. vaginatus Turcz. (P. x bottnicus Hagstr.). As we have recently published a detailed account of these populations (Preston et al. 1998) they are not considered further in this paper. DISTRIBUTION OF P. x SUECICUS IN THE BRITISH ISLES The known sites for P. x suecicus are listed below, with the first record(s), any later records of particular interest and records made by us since 1986. The records are arranged by vice-county and site, with a site being taken as a single river or stream or as a lake together with its associated inflow and outflow streams and ditches. We have cited appropriate literature references or herbarium specimens to support all records; unpublished determinations by Dandy are cited from the “Dandy index’, the card index of herbarium specimens compiled by Dandy and now at BM. Almost all records of P. x suecicus made since 1940 are supported by specimens determined by Dandy & Taylor or us; the main exceptions are literature references to sites in the Outer Hebrides in the works of J. W. Heslop Harrison and his colleagues, which we summarise after the accepted records for the vice-county. The records of P. x suecicus accepted below are mapped in Fig. 1. The identity of some of the populations has been confirmed by isozyme analysis. Isozyme studics by Hollingsworth et al. (1996a, b) revealed consistent differences between the parent species in four enzyme systems, with P. x suecicus showing the additive inheritance which would be expected of the hybrid between them. Sites are marked with an double asterisk (**) if the isozyme results are described in detail by Hollingsworth et al. (1996a) or by a single asterisk (*) if the morphological identity is supported by unpublished results based on the diagnostic AAT and/or IDH enzyme systems. 332 C. D. PRESTON, P. M. HOLLINGSWORTH AND R. J. GORNALL MID-W. YORKS. (V.C. 64) **River Wharfe: near Pool, SE/2.4, 1868, F. A. Lees (Dandy & Taylor 1946); Linton Bridge, SE/3.4, 1880, F. A. Lees (Dandy & Taylor 1946); Wetherby, SE/4.4, 1881, J. Jackson (Dandy & Taylor 1946); Arthington, SE/2.4, undated but probably collected before 1890, J. Abbot (Dandy & Taylor 1946); Arthington, Leathley & Weeton, SE/2.4, Linton, Harewood & Netherby, SE/3.4, Wetherby, SE/4.4, Ozendyke, SE/5.3, and Ulleskelf, SE/5.4, 1940-1945, G. Taylor (Dandy & Taylor 1946); Harewood Bridge, SE/31.46, East Keswick, SE/34.46, Linton Bridge, SE/38.46, and Linton, SE/39.47, 1989-1996, Mrs P. P. Abbott, P.M.H. & C.D.P. (CGE, E, LTR, NMW). River Ure: Sharow, SE/3.7, 1875, G. Nicholson (Dandy index); Ripon, SE/3.7, 1881, H. H. Slater (Dandy & Taylor 1946); Littlethorpe, SE/3.6, Nunwick, SE/3.7, and Ripon, SE/3.7, 1940-1942, G. Taylor (Dandy & Taylor 1946); near Ripon, SE/3.7, 1950, U. K. Duncan & C. M. Rob (Dandy index). Tnbutary of River Ure, Westwick, SE/3.6, 1943, G. Taylor (Dandy & Taylor 1946). N.W. YORKS. (V.C. 65) **River Ure: West Tanfield, SE/2.7, Masham, SE/2.8, Langthorpe, SE/3.6 and Norton Conyers, SE/3.7, 1940-1945, G. Taylor (Dandy & Taylor 1946); West Tanfield, SE/26.78 and Masham, SE/22.81, 1988-1996, P.M.H. & C.D.P. (CGE, LTR, NMW, RNG). [CHEVIOT (V.C. 68) AND BERWICKS. (V.C. 81) Records from these vice-counties are based on material which we have identified as P. x bottnicus (Preston et al. 1998). Both P. filiformis and P. pectinatus are found in Coldingham Loch, v.c. 81, but we have been unable to find the hybrid there although the locality appears suitable. | [MIDLOTHIAN (V.C. 83) | The record from Duddingston Loch published in Watsonia 15: 138 (1984) was retracted in Watsonia 17: 481 (1989) as it was based on a specimen which is indistinguishable from P. pectinatus. | FIFE (V.C. 85) **T_ och Fitty, NT/12.91, 1992-1994, P.M.H. & C.D.P. (BM, CGE, E, LTR). *Cameron Reservoir, NO/477.113, 1998, J. M. Croft, R.J.G. & C.D.P. (BM, CGE, E, LTR). N. ABERDEEN (V.C. 93) Canal, St Fergus, NK/0.5, 1876, J. H. Walker (Dandy & Taylor 1940): the canal is now disused and dry for most of its length and P.M.H., C.D.P. & D. Welch were unable to find any Potamogeton species except P. natans in the very shallow pools remaining near Inverugie, NK/09.48, in August 1994. *Loch of Strathbeg, Starnakeppie, NK/083.585, 1994, P.M.H., C.D.P. & D. Welch (CGE, E). MORAY (V.C. 95) Innes Canal, Urquhart, NJ/2.6, 1946, G. Taylor (Dandy & Taylor 1946); not refound by C.D.P. & P.M.H., 1994. **River Lossie: Calcots, NJ/2.6, 1967 & 1972, M. McCallum Webster (Dandy index, cf McCallum Webster 1978); Bndge of Calcots, NJ/254.638, and Arthur’s Bridge, NJ/253.672, 1994, P.M.H. & C.D.P. (CGE, E, LTR, NMW). Loch Spynie, NJ/2.6, 1972, M. McCallum Webster (Dandy index, cf McCallum Webster 1978); NJ/237.663, 1994, P.M.H. & C.D.P. (CGE, E). S. EBUDES (V.C. 102) Loch Fada, Colonsay, NR/3.9, 1908, M. McNeill (Dandy & Taylor 1940). MID EBUDES (V.C. 103) ** Abhainn a’Bheidhe (the stream from Loch a’Phuill to Balephuil Bay), Tiree, NL/9.4, 1940, W. A. Clark (Dandy index; cf Heslop Harrison et al. 1941; Heslop Harmison 1949); 1968, U. K. Duncan (Dandy index; cf Duncan 1969); 1989-1997, P.M.H., C.D.P. & N. F. Stewart (CGE, E, LTR). Ditch N.W. of Loch a’Phuill, Tiree, NL/953.422, 1990, D. A. Pearman, det. C.D.P. (CGE, E). **Loch a’Phuill, Tiree, NL/9.4, 1993, R. N. Evans & P.M.H. (LTR). DISTRIBUTION OF POTAMOGETON x SUECICUS 333 ** An Fhaodhail, Tiree, NM/0.4, 1897, S. M. Macvicar (Dandy & Taylor 1940); in main stream, a large backwater (Poll Orisgal) and nearby pools, 1989-1997, P.M.H., C.D.P. & N. F. Stewart (CGE, E, LTR). OUTER HEBRIDES (V.C. 110) Loch nam Budh, Monach Island, NF/63.61, 1949, F. H. Perring, det. C.D.P. (CGE, cf Preston in press). *Loch Stlligarry, S. Uist, NF/766.379, 1995, P.M.H. & C.D.P. (BM, CGE, LTR). *Inflow stream to Loch Stilligarry, S. Uist, NF/767.380, 1995, P.M.H. & C.D.P. (CGE). **West Loch Ollay, S. Uist, NF/738.324, 1994-1995, P.M.H. & C.D.P. (CGE, E, LTR). *Loch an Duin Bhig, S. Uist, NF/759.468 & 760.470, 1995, P.M.H. & C.D.P. (BM, CGE, E, LTR). **]_och na Liana Moire, Benbecula, NF/76.53, 1940, W. A. Clark (Dandy & Taylor 1941; cf Heslop Harrison 1941, 1949; Heslop Harnson & Clark 1941); 1987, C.D.P., N. F. Stewart et al. (CGE, cf Preston 1991); 1994-1995, P.M.H. & C.D.P. (CGE, LTR). *Ditch between Loch na Liana Moire and Loch Torcusay, Benbecula, NF/763530, 1995, P.M.H. & C.D.P. (CGE). *Loch Torcusay, Benbecula, NF/761.532, 1995, P.M.H. & C.D.P. (BM, CGE, E). **]_ och Fada, Benbecula, NF/773.518, 1994-1995, P.M.H. & C.D.P. (BM, CGE, LTR). Loch near Borve Castle [this may be Loch a’Chinn Uacraich], Benbecula, NF/7.5, 1940, W. A. Clark (Dandy & Taylor 1941; cf Heslop Harrison 1941, 1949; Heslop Harrison & Clark 1941). **T och a’Chinn Uacraich, Benbecula, NF/767.510, 1994-1995, P.M.H. & C.D.P. (CGE, E, LTR). Lochan near Uachdar, Benbecula, NF/7.5 or 8.5, 1940, W. A. Clark (Dandy & Taylor 1941; cf Heslop Harrison 1941; Heslop Harrison & Clark 1941). The record from a lochan near Gramisdale, NF/8.5, cited by Heslop Harrison (1949) may refer to the same site. *Loch na Paisg, Baleshare, NF/786.618, 1995, P.M.H. & C.D.P. (BM, CGE). *Loch Mor, Baleshare, NF/789.621, 1995, P.M.H. & C.D.P. (CGE). *Loch Sandary, N. Uist, NF/734.684, 1995, P.M.H. & C.D.P. (CGE, E, LTR). **T_ och Grogary, N. Uist, NF/71.70 & 71.71, 1994-1995, P.M.H. & C.D.P. (CGE, E, LTR). *Loch Scarie, N. Uist, NF/716.704, 1995, P.M.H. & C.D.P. (BM, CGE, E, LTR). *Loch a’Chaolais, N. Uist, NF/897.780, 1995, P.M.H. & C.D.P. (BM, CGE, E). *Loch Bhruist, Berneray, NF/9.8, 1938, J. W. Campbell & 1939, A. J. Wilmott (Dandy & Taylor 1940); NF/915.822 & 920.829, 1995, P.M.H. & C.D.P. (CGE, E, LTR). *Outflow stream at S. end of Loch Bhruist, Berneray, NF/914.820, 1995, P.M.H. & C.D.P. (BM, CGE). Little Loch Borve, Berneray, NF/91.81, 1939, W. A. Clark (Dandy & Taylor 1940; cf Clark & Heslop Harrison 1940; Heslop Harrison 1941, 1949); 1995, P.M.H. & C.D.P. (BM, CGE). *Outflow stream S. of Little Loch Borve, Berneray, NF/911.814, 1995, P.M.H. & C.D.P. (CGE). Specimens at BM and CGE collected by A. J. Wilmott from Loch na Doirlinn, Barra, in 1938 (380718La) may be P. X suecicus but are inadequate for certain identification (Preston in press). We have not traced voucher specimens to support the records of P. x suecicus from Loch Bornish, Loch Hallan and lochs near Stoneybridge, all on S. Uist (Heslop Harrison 1949; Heslop Harrison & Clark 1942; Heslop Harmison et al. 1942) and Loch Cistavat, S. Harris (Heslop Harrison & Heslop Harrison 1950). ORKNEY (V.C. 111) Loch of the Riv, Sanday, HY/68.46, 1994, N. F. Stewart, det. C.D.P. (CGE, E, cf Preston & Stewart 1995). Loch of Langamay, Sanday, HY/74.44, 1963, E. R. Bullard, det. C.D.P. (BM, cf Preston in press); 1986, E. Charter, det. C.D.P. (CGE, NCCE, cf Preston & Stewart 1995); 1994, N. F. Stewart, det. C.D.P. (CGE, E, LTR, cf Preston & Stewart 1995). Loch of Rummie, Sanday, HY/75.44, 1920, H. H. Johnston, det. C.D.P. (E, cf. Preston in press); 1994, N. F. Stewart, det. C.D.P. (BM, CGE, E, cf Preston & Stewart 1995). Loch Gretchen, North Ronaldsay, HY/74.52, 1920, H. H. Johnston, det. C.D.P. (E, cf. Preston in press). SHETLAND (V.C. 112) *Loch of Clickimin, HU/4.4, 1980, R. C. Palmer, det. C.D.P. (SLBI, herb. R.C.P., cf Preston in press); HU/46.41, 1996, P.M.H. & C.D.P. (BM, CGE, E). 334 C. D. PRESTON, P. M. HOLLINGSWORTH AND R. J. GORNALL S. KERRY (V.C. H1) *Lough Gill, V/61.14, 1993, N. F. & R. J. Stewart; V/61.13 & 61.14, 1994, R. FitzGerald & C.D.P. (BM, CGE, DBN, LTR). ROSCOMMON (V.C. H25) *Callow Lough, 1-5 km S. of Cuil Bridge, M/70.96, 1998, A. B. Carter, D. C. F. Cotton, A. Hill, N. Raftery & C.D.P. (CGE, DBN). E. MAYO (V.C. H26) Glore River, Kiltamagh, M/38.90, 1994, C.D.P. (CGE, DBN). Similar plants grew upstream at gnd reference M/38.89 and M/40.89 but were not collected as in the field the plant was thought to be P. pectinatus. W. DONEGAL (V.C. H35) Rosapenna, C/11.38, 1989-1990, C.D.P. & N. F. Stewart (BEL, BM, CGE, DBN, LTR, cf Preston & Stewart 1994). TABLE 1. SPECIES ASSOCIATED WITH POTAMOGETON x SUECICUS IN RIVERS, STREAMS AND LAKES Rivers Streams Lakes (1) Lakes (2) Potamogeton X suecicus V (5-10) V (5-10) ViG,9) V (4-10) Myriophyllum spicatum I (4) III (1-7) I (1-4) III (1-7) Potamogeton crispus II (2-5) I (6) I (7) I (1) Potamogeton perfoliatus I (4) GD) II (1-2) Potamogeton natans II (3-6) II (2—S) Ill (1-5) Lemna minor Ir(d—2) I (2-3) Elodea canadensis 1 (6) I (1-5) Fontinalis antipyretica I (4) (GD) Eleocharis palustris III (1-3) III (1-6) II (1-4) Potamogeton filiformis II (2-3) I (1-6) II (6-8) Potamogeton friesii I (4) II (1-2) I (1-4) Hippuris vulgaris I (3) I (2) I (3-5) Potamogeton X nitens I (2) I (1) (G=6) Ranunculus baudotii I (2) I (6) I (5) Equisetum fluviatile i) II (1-3) Chara aspera 106) III (1-8) Chara vulgaris If] (1-3) If (1-5) Chara contraria II (6) ] (1-5) Potamogeton pusillus 1 (3) II (1-7) Littorella uniflora IT (1-8) 1 (2) Callitriche hermaphroditica EOL) ] (2-4) Potamogeton pectinatus 1G) I (6) Agrostis stolonifera III (1) Sparganium erectum II (1-5) Persicaria amphibia II (4-7) Chara hispida I 3-7) Chara virgata I (1-3) Phragmites australis I (2-3) Baldellia ranunculoides K@S) Carex rostrata I (1-2) Lemna trisulca II (1-2) Potamogeton gramineus I (1-5) Zannichellia palustris I (2-4) DISTRIBUTION OF POTAMOGETON x SUECICUS TABLE 1. CONTINUED Rivers Streams Lakes (1) Lakes (2) No. quadrats 9 1! 10 14 No. vascular plants & bryophytes mean 2-3 5-0 4-7 4-9 (range) (1-4) (2-8) (1-8) (2-8) No. charophytes mean 0 0-9 1-2 0 (range) (1-2) (1-4) % bare ground mean 16 9 5 15) (range) (0-45) (0-15) (0-15) (0-50) Water depth (cm) mean 41 18 20 38 (range) (15-80) (15-25) (10-45) (12-70) Substrate type (% quadrats) rocks 40 0 0 0 stones/gravel 30 0 0 5 sand 20 50 70 30 silt/mud 10 50 30 65 The roman numerals indicate the percentage of quadrats in which the taxon was recorded in each habitat: I, 1—20%; II, 21-40%; III, 41-60%; IV, 61-80% and V, 81-100%. The figures in brackets indicate the range of Domin cover-abundance values recorded. The lake quadrats are subdivided into two groups based primarily on the presence or absence of charophytes. The following taxa were recorded in a single quadrat: Callitriche hamulata, Sparganium emersum (rivers); Caltha palustris, Mentha aquatica, Ranunculus trichophyllus, Rorippa X sterilis, Veronica anagallis-aquatica (streams); Myriophyllum alterniflorum, Potamogeton xX billupsii (lakes 1); Potamogeton rutilus, P. x zizti (lakes 2). HABITAT OF P. x SUECICUS IN BRITAIN = Between 1994 and 1997 we visited 21 of the 28 localities in which P. x suecicus has been recorded in Britain recently, and recorded its habitat in forty 4m’ quadrats. At each site one or more quadrat was recorded in stand(s) of vegetation where the cover of the hybrid was greatest. We assessed the cover-abundance of vascular plants, bryophytes and charophytes in each quadrat and noted details of substrate and water depth. The quadrats covered the range of water bodies from which the hybrid has been reported, including the River Lossie (v.c. 95), River Wharfe (v.c. 64) and River Ure (v.c. 65), the streams Abhainn a’ Bheidhe and An Fhaodhail on Tiree (v.c. 103) and the outflow of Little Loch Borve, Berneray (v.c. 110), the ditch between Loch Torcusay and Loch Liana Moire, Benbecula (v.c. 110), and 15 lakes in Fife (Loch Fitty), Moray (Loch Spynie) and the Outer Hebrides (on Benbecula, Berneray, N. Uist and S. Uist). The number of sites covered should ensure that the quadrat data characterise the range of vegetation in which the hybrid occurs, although stands in deep water may have been overlooked. The vegetation in the quadrats is summarised in Table 1. The quadrats from rivers, streams and ditches, and lakes are separated. The quadrats from lakes are presented in two groups, split primarily on the presence or absence of charophytes in the quadrat. In rivers P. X suecicus is often present as robust plants growing in large, dominant stands. Very large, dense stands of P. x suecicus include those in the River Wharfe near Harewood Bridge (SE/31.46), East Keswick (SE/36.45) and Linton Bridge (SE/38.46) and the River Lossie near Bridge of Calcots (NJ/25.63) and Arthur’s Bridge (NJ/25.67). Large stands of P. x suecicus are less frequent in the River Ure but the hybrid may often be found in abundance by the bridge at West Tanfield (SE/26.78), where it was photographed on 6 August 1945 (Dandy & Taylor 1946) and has therefore survived for over SO years. In rivers P. x suecicus 1s rooted in a substrate of rocks, stones, 336 C. D. PRESTON, P. M. HOLLINGSWORTH AND R. J. GORNALL gravel or (on stretches of the River Lossie) pure sand. Where the substrate is rocky, plants of P. x suecicus are often rooted under boulders, stones or the masonry of bridge supports. Although the water flow in the P. X suecicus rivers is relatively rapid, patches of P. x suecicus may be so large that they impede the flow of water when they reach the surface, allowing a few fronds of Lemna minor to settle amongst them. Large stands may flower freely: flowering P. x suecicus in the River Lossie in August 1994 bore 9-42 inflorescences in sample areas of 400 cm’, equivalent to 225-1050 m’. These totals include inflorescences at anthesis and others which were decaying; some were on the surface of the water and others buried amongst the submerged foliage. The river quadrats in Table 1 are species-poor. The vegetation in these quadrats is similar to that in which the related hybrid P. x bottnicus grows in the River Till and River Tweed (Hollingsworth et al. 1998). Although no Ranunculus species were recorded in the quadrats with P. x suecicus, R. penicillatus subsp. pseudofluitans and R. x bachii grow in the same stretches of the Wharfe as P. x suecicus. The streams and ditches in which P. X suecicus was recorded in quadrats are shallow and flow slowly and gently over substrates of sand or silt. P. X suecicus may span the entire width of the narrow channel of both Abhainn a’Bheidhe and An Fhaodhail in Tiree. Heslop Harrison (1949) reported that P. x suecicus grew at Abhainn a’ Bheidhe in “dense masses for considerable stretches of the stream’, still an apt description of its abundance in 1997. The stream quadrats are more species-rich than those recorded in rivers, and in addition to submerged species they include emergents such as Eleocharis palustris, Sparganium erectum and a few shoots of the normally terrestrial Agrostis stolonifera which extend into the shallow water of An Fhaodhail. The vegetation is usually more or less closed and there 1s little bare ground. In some lakes P. X suecicus may occur 1n large, dominant stands. The largest stands we have seen in lakes have been in the very shallow bay at the northern end of Loch Torcusay (NF/76.53) where the water is only 10 cm deep and is completely dominated over an area of many square metres by P. X suecicus and associated charophytes (Chara contraria, C. hispida, C. virgata and C. vulgaris), and in the north-west arm of Loch an Duin Bhig (NF/76.47) which is covered in P. X suecicus. Equally dense but smaller stands may be found in other lakes, such as Loch Fitty (NT/12.91) and Loch Sandary (NF/73.68). In other sites P. X suecicus may occur only as scattered plants in more Open communities over stones and boulders, or amongst other macrophytes such as Hippuris vulgaris, Littorella uniflora, Persicaria amphibia, Potamogeton filiformis, P. x nitens, P. pectinatus or Zannichellia palustris. The quadrats where P. X suecicus was recorded with charophytes (Table 1) tend to be in shallower water than those which lack charophytes, and are more often found with sand as a substrate than silt. Eleocharis palustris, Potamogeton natans and Chara aspera are the most frequent associates in these quadrats and Persicaria amphibia is confined to this group. Charophyte cover is often high and there is little bare ground. Eleocharis palustris is less frequent in the quadrats without charophytes and neither Persicaria amphibia nor Potamogeton natans are recorded. However, Eguisetum fluviatile 1s present and Myriophyllum spicatum and Potamogeton pusillus are more frequent in these quadrats. P. filiformis and P. pectinatus occur with P. X suecicus in both groups of quadrats. HABITAT OF P. x SUECICUS IN IRELAND P. X suecicus grows in lakes at three of its four known Irish sites. Two of these lakes are similar to sites in Scotland where the hybrid occurs in charophyte-rich communities. In Donegal (v.c. H35) it is found in a shallow lake in the calcareous sand dunes at Rosapenna. This site, which is described by Preston & Stewart (1994), is fringed by Eleocharis palustris, Littorella uniflora and scattered Persicaria amphibia. In the open water charophytes are abundant and in addition to P. x suecicus the macrophytes include Apium inundatum, Potamogeton natans and Ranunculus trichophyllus. The second Irish site is a large coastal lake, Lough Gill (v.c. H1). Here P. x suecicus is locally frequent at the south-east edge of the lake, growing in shallow water 12-15 cm deep over a substrate of silt mixed with stones and sand. Associated species include Potamogeton filiformis, P. pusillus, Chara contraria and C. curta. By contrast, in Roscommon (v.c. H25) P. x suecicus has been found in water 30-50 cm deep over a stony substrate at the edge of Callow Lough, the southernmost lough in the Lough Gara complex. Here it grows as scattered plants or, in more sheltered bays, as somewhat larger patches, with few associated species. DISTRIBUTION OF POTAMOGETON x SUECICUS 337 TABLE 2. SITES WHERE POTAMOGETON x SUECICUS MAY GROW IN THE ABSENCE OF ONE OR BOTH PARENTS, P. FILIFORMIS (F) AND P. PECTINATUS (P) Site Parents Notes present River Wharfe, v.c. 64 Je South of current distribution of F River Ure, v.c. 64, 65 Je South of current distribution of F Canal, St Fergus, v.c. 93 FR Both taxa known only from specimens dated 1876 Loch of Strathbeg, v.c. 93 IP F recorded by Trail (1901a,b) but never confirmed Innes Canal, Urquhart, v.c. 95 Neither Hybrid known only from specimen dated 1946 River Lossie, v.c. 95 Neither Loch Spynie, v.c. 95 IP P collected regularly since 1831; F never found Loch Fada, v.c. 102 Ie F recorded by McNeill (1910, p. 78); no specimen seen and species has not been reported from the island again (Clarke & Clarke 1991) Loch nam Budh, v.c. 110 Neither Specimen determined as F by Dandy and reported as such by Perring & Randall (1972) is the hybrid Loch an Duin Bhig, v.c. 110 Je F recorded by Royal Botanic Garden, Edinburgh (1983); no specimen seen Loch Fada, v.c. 110 P Loch a’Chinn Uacraich, v.c. 110 P F recorded by Royal Botanic Garden, Edinburgh (1983); . no specimen seen Loch Mor, v.c. 110 P Loch a’Chaolais, v.c. 110 Neither Small site thoroughly surveyed in 1995 Little Loch Borve, v.c. 110 F Loch of the Riv, v.c. 111 F Loch of Langamay, v.c. 111 Je Specimen collected in 1963 and determined by Dandy as P is the hybrid (Preston in press) Loch of Rummie, v.c. 111 F Specimen reported by Johnston (1922) as F but determined by Dandy & Taylor as P is the hybrid (Preston in press) Glore River, v.c. H26 Neither Rosapenna, v.c. H35 F F collected in 1939 by Praeger but absent when this small site was thoroughly surveyed in 1989 & 1990 The remaining Irish site, the Glore River (v.c. H26), is a relatively shallow stream flowing over a substrate of silt and stones. P. X suecicus 1s recorded in patches up to 3 metres long, growing in water 40-50 cm deep with Apium nodiflorum, Elodea canadensis, Myriophyllum alterniflorum and the rare hybrid Potamogeton xX lanceolatus. This stream has no close parallel to any of the other known sites for P. x suecicus in Britain, and perhaps provides a habitat which is intermediate between the large rivers of northern England and Scotland and the slowly flowing streams of the Hebnides. OCCURRENCE OF P. x SUECICUS IN RELATION TO ITS PARENTS In discussing the relationship between the distribution of P. xX suecicus and its parents, three questions arise. What 1s the probability of finding P. x suecicus at sites where both Potamogeton filiformis and P. pectinatus occur? Does the hybrid occur at sites where one or both parents are absent? And is there any differentiation between the habitat of the three taxa when two or three grow together? It is not easy to assess the frequency of the hybrid at sites where both parents occur as there are few areas in which the hybrid has been searched for systematically. In the Outer Hebrides in 1995 we visited 17 lochs which appeared to be likely sites for the hybrid, comprising 3 where P. x 338 C. D. PRESTON, P. M. HOLLINGSWORTH AND R. J. GORNALL suecicus had already been recorded before we began our studies and 14 from which one or both parents had been recorded or seemed likely to occur. P. X suecicus grew in 15 of these 17 sites, the exceptions being one loch where we were only able to detect the parents in the limited area we were able to search and one small loch near the sea where P. pectinatus was the only one of the three taxa present. This is clear evidence to support Heslop Harnson & Clark’s (1941) view that on Benbecula the hybrid “seems to occur wherever the parent species clash”. However, P. x suecicus may be exceptionally frequent in the Outer Hebrides. In Shetland, where P. filiformis is much more frequent than P. pectinatus, we have visited all seven sites from which Scott & Palmer (1987) cite confirmed records of P. pectinatus. Despite the fact that P. filiformis grows in six of these seven sites, we detected P. X suecicus in only one locality. The sites where P. X suecicus appears to grow in the absence of one or both parents are summarised in Table 2. The identification of such sites is not always straightforward. At some sites there are field or literature records of one of the parents made by recorders who did not report the hybrid; if these records are not supported by herbarium material it is not possible to say whether they are correct. At other sites the absence of the parents may be due to inadequate survey. Nevertheless, there are several localities where the presence of P. x suecicus in the absence of one or both parents is well-established. Three of the sites are rivers: the River Lossie, where neither parent is found, and the Rivers Wharfe and Ure, where only P. pectinatus occurs. The recently discovered population in the Glore River may also fall into this category. P. filformis does not usually occur in large rivers, and its absence from these sites may be explicable on ecological grounds. The Yorkshire rivers are also south of the current range of P. filiformis. Another habitat where P. x suecicus occurs in the absence of one or both parents is in shallow coastal lakes over sand. At Loch a’Chaolais and Rosapenna the hybrid is currently present and locally abundant in the absence of both parents. Ecologically these sites have much in common; both sites are fringed by Eleocharis palustris swamp and Persicaria amphibia, Potamogeton natans and Chara aspera or the closely related C. curta grow in the water. There is reliable evidence for the former presence of one of the parents at Rosapenna, as R. L. Praeger collected fruiting material of P. filiformis there in 1939 (DBN). The Orkney sites for P. x suecicus are also shallow lakes over calcareous sand where the hybnd grows with Chara aspera and C. curta; it 1s accompanied here by P. filiformis but there is no reliable record of P. pectinatus from these sites. The presence of “orphaned” P. x suecicus in larger lakes has still to be established with certainty but it is interesting that the hybrid grows in Loch Spynie, where many collectors from 1831 onwards have gathered P. pectinatus but where P. filiformis has not been recorded. The Loch of Strathbeg is also a possibility: P. pectinatus has been recorded since 1883 but the only record of P. filiformis (Trail 1901a, b) lacks a supporting specimen. At large and complex sites where the hybrid and both parents are present, the taxa are often concentrated in different areas. At Loch a’Phuill P. filiformis grows on the shallow sandy flats around the edge and P. pectinatus in deeper water. P. xX suecicus occurs in the loch but it is particularly abundant in the shallow and slowly flowing outflow stream (Abhainn a’ Bheidhe) which runs from this loch through sand dunes to the sea. P. filiformis grows in this stream (though itis less abundant than the hybrid) but P. pectinatus does not. At An Fhaodhail, Tiree, the shallow river is dominated by P. x suecicus; P. pectinatus is represented by a few individuals scattered sporadically amongst the hybrid and P. filiformis grows in nearby pools (Hollingsworth et al. 1996a). At less complex sites the distinctions between the taxa are less obvious. P. filiformis tends to be a plant of shallow water and although it may grow with P. pectinatus, the latter attains maximum luxuriance in deeper, less turbulent water (van Wijk 1988). Unlike P. filiformis, which may be found in water only a few centimetres deep and in sites which dry out completely when water levels arc low, P. x suecicus is not found in very shallow water. It does, however, grow in slightly deeper water well within the habitat range of P. filiformis and it is more frequently found with that species than with P. pectinatus (Table 1). It would, however, be even more easily overlooked in deeper water than it is when it grows in the shallows. At Loch Fitty P. filiformis tends to grow on coarse sand or gravel, P. pectinatus on fine silt and P. x suecicus gravel mixed with silt, although this may simply reflect differences in the water depth and exposure of the sites favoured by the three taxa. In summary, the habitat of P. x suecicus in lakes is intermediate between that of its parents. It tends to be most abundant in water at the deeper end of the range characteristic of P. filiformis, Dut DISTRIBUTION OF POTAMOGETON x SUECICUS 339 it rarely extends into the still deeper water where P. pectinatus reaches maximum luxuriance. It may grow in the absence of both parents in shallow coastal lakes over sand. In rivers it grows in shallow, fairly rapidly flowing water over stones, gravel or sand, a habitat where neither parent is found. A more detailed insight into the habitat of the hybrid in relation to its parents might be obtained by more intensive studies at sites like Loch Fitty where all three taxa occur. DISCUSSION RECOGNITION AND RECORDING HISTORY Plants of P. X suecicus can easily be confused with P. pectinatus as the resemblance to that species is obvious in the field whereas the influence of P. filiformis is often apparent only when plants are carefully examined under the binocular microscope. Many specimens of P. X suecicus have initially been identified (at least in the field) as P. pectinatus: these include those collected by G. Taylor from the River Wharfe in 1940, the first specimens collected from Shetland in 1980, plants gathered by C.D.P. & N. F. Stewart in Co. Donegal in 1989 and material collected by P.M.H. at Loch Fada and Loch a’Chinn Uacraich, Outer Hebrides, in 1994. P. x suecicus has also been overlooked (rather than mistaken for its parents) in Scotland and Ireland, perhaps because fruiting plants of P. filiformis and P. pectinatus may often be found without difficulty, so that botanists can record both species with certainty and therefore do not feel obliged to examine vegetative material. The ease with which P. x suecicus can escape detection is well illustrated in the Outer Hebrides. The hybrid is now known to be widespread in species-rich machair lochs which have attracted the attention of many individual botanists and ecologists and some survey teams in recent years (Preston 1991). Rare and critical taxa such as Potamogeton x billupsii, P. x nitens, P. rutilus and P. x sparganifolius were collected between 1960 and 1985 from lochs in which P. Xx suecicus is now known to occur, but P. X suecicus itself was never recorded during this period. The fact that Potamogeton X suecicus tends to be recorded only by botanists who are familiar with its appearance, and is overlooked even by others who are specifically recording in aquatic habitats, explains the numerous records in the 1940s, when both Heslop Harrison’s team from Newcastle and George Taylor were actively engaged in fieldwork, followed by the subsequent falling-off of new records. The hybrid is almost certainly still under-recorded. One would expect to find it elsewhere in eastern Scotland (e.g. in Angus) and at other sites in the west. There are many potential sites for the hybrid in the Outer Hebrides, for example, which we have not had an Opportunity to visit. Four sites have been discovered in Ireland in the last decade and there are many more places in the north and west where it might occur. We have improved our ability to recognise P. x suecicus only by collecting material in the field, examining it from a morphological or isozyme perspective and then returning to the field to reassess the populations in the light of these detailed studies. On our return from visits to N.E. Scotland, the Hebrides and Shetland C.D.P. has examined the morphology of our collections and P.M.H. has looked at the isozymes. We have then compared identifications which were reached without knowledge of the other person’s view. We have always been in agreement, though sometimes the isozyme results have provided welcome confirmation of a tentative identification of fragmentary or aged material based on morphology, or vice versa. However, an identification based on morphology often requires flowering material, or a supply of vegetative material which is large enough to allow the dissection of numerous young leaf sheaths. The hybrid is more likely to be detected by a thorough examination of populations in the field than by the collection of a few herbarium specimens for later examination or determination by others. Recognition of the hybrid using morphological criteria relies on proving the presence of characters derived from both parents, and it is helpful to demonstrate that the plant is sterile. Herbarium material often provides an inadequate representation of the habit of the plant and a very small number of leaf sheaths for dissection. Flowers may not be available, and even if a flowering plant lacks fruit there is usually no evidence to indicate whether this was typical of the population from which it was collected. Even J. E. Dandy and G. Taylor, the foremost authonities on the genus, identified material which we now believe to be P. X suwecicus as one or other parent (Preston in press). 340 C. D. PRESTON, P. M. HOLLINGSWORTH AND R. J. GORNALL DOES P. x SUECICUS BACKCROSS WITH EITHER PARENT? Heslop Harrison & Clark (1942), in commenting on the application of the name P. x suecicus to the hybrid plant from South Uist which “grows in the Stoneybridge lochs, as elsewhere, in forms displaying a great variation range”’ said that “in our opinion, it is wrong to include these in one ragbag under the name P. suecicus because they differ phenotypically and genotypically. Clearly, F, hybrids, backcrosses and segregates of F, and later generations are concerned; a common name cannot be forced to cover plants ranging from “‘almost” P. filiformis to “almost” P. pectinatus. It would, at present, be better to label them P. filiformis x P. pectinatus, and to leave them until further study and experiment have clarified the position.” This description suggests that a complex hybrid swarm is present in the Hebrides. We know of no evidence to support this suggestion. The plants of P. x suecicus in those localities where we have studied them in detail show little variation. The hybrid does vary in morphology from population to population, and some of these variants are closer to P. pectinatus than others, but this pattern of variation is much more easily explained by phenotypic response to different habitat factors or by the presence in different sites of a first generation (F,) hybrid of different genetic origin than it is by invoking the possibility of backcrossing with the parents. The isozyme studies of eleven populations reported by Hollingsworth et al. (1996a) confirm this pattern of variation: no variation was detected at six sites and only two isozyme phenotypes were detected at a further four. This suggests that there may have been only one or two clones present at ten of the localities (although isozyme data are based on a small proportion of the genome and provide an estimate of the minimum number of clones present). An Fhaodhail on Tiree, where six isozyme phenotypes were detected, was the only exception. All isozyme phenotypes, including those from An Fhaodhail, were consistent with the assumption that the plants were F, hybrids, and there was no evidence for the disruption of additive inheritance of species-specific markers which would be expected if backcrossing had taken place. SIGNIFICANCE OF STERILE POTAMOGETON HYBRIDS P. X suecicus is one of eight widespread Potamogeton hybnids in Britain and Ireland identified by Preston (1995) and mapped by Preston & Croft (1997). Like most of these hybnids, its distribution is neither completely independent of the parents nor completely determined by the sites where they currently occur. Despite the fact that P. x suecicus was not recognised in Britain until 1940, it is known to have been present in three sites for at least 100 years (River Wharfe, River Ure, An Fhaodhail) and in a further five for at least 50 years (Abhainn a’Bheidhe, Loch na Liana Moire, Loch Bhruist, Little Loch Borve and Loch of Rummie). In sites where at least one parent is absent the presence over this period is almost certainly caused by the persistence of particular clones; if both parents are present one cannot conclude that individual clones are long-lived as the hybrid may have arisen by repeated hybridisation. Another notable feature that P. x suecicus shares with many of the widespread hybrids is that at some sites it is a significant feature of the aquatic vegetation, and may even be present in dominant stands. The further study of the ecology of these sterile hybrids in relation to their fertile parents might throw considerable light on the reproductive biology, dispersal and ecology of aquatic plants. ACKNOWLEDGMENTS We are particularly grateful to N. F. Stewart for his help on initial visits to P. x suecicus sites, and for later contributing records from his own fieldwork; he also identitied the charophytes cited in this paper and provided helpful comments on an earlier draft. We also thank Mrs P. P. Abbott, Miss A. B. Carter, Dr D. C. F. Cotton, Mrs J. M. Croft, R. N. Evans, Lady Rosemary FitzGerald, A. Hill, D. A. Pearman, C. S. & Mrs M. T. Preston, N. Raftery and D. Welch for helping us with fieldwork, Ms E. Charter, Dr O. Lassiére & R. C. Palmer for sending herbarium material and D. R. McKean & R. Vickery for help on visits to herbaria. The map was plotted using the DMAP program written by Dr A. J. Morton. This work was supported in part by a N.E.R.C. CASE studentship and by a contribution towards fieldwork expenses in Shetland from the B.S.B.I. Bequest Fund. DISTRIBUTION OF POTAMOGETON x SUECICUS 34] REFERENCES Bance, H. M. (1946). A comparative account of the structure of Potamogeton filiformis Pers. and P. pectinatus L. in relation to the identity of a supposed hybrid of these species. Transactions of the Botanical Society of Edinburgh 34: 361-367. Ciark, W. A. (1943). Pondweeds from North Uist (v.c. 110), with a special consideration of Potamogeton rutilus Wolfg. and a new hybrid. Proceedings of the University of Durham Philosophical Society 10: 368-373. CLarK, W. A. & HeEsLop Harrison, J. (1940). Noteworthy plants from Great and Little Bernera (Lewis), Pabbay and Berneray (Harris), and the Uig district of Lewis. Proceedings of the University of Durham Philosophical Society 10: 214-221. CrarkE, P. M. & Crarke, J. (1991). The flowering plants of Colonsay and Oransay. Privately published. Danpy, J. E. (1975). Potamogeton L., in Stace, C. A., ed., Hybridization and the Flora of the British Isles, pp. 444-459. Academic Press, London. Danpy, J. E. & Tay or, G. (1940). Studies of British Potamogetons. — XIV. Potamogeton in the Hebrides (Vice-county 110). Journal of botany 78: 139-147. Danpy, J. E. & Taytor, G. (1941). Studies of British Potamogetons. — XV. Further records of Potamogeton from the Hebrides. Journal of botany 79: 97-101. Danpy, J. E. & Taytor, G. (1946). An account of x Potamogeton suecicus Richt. in Yorkshire and the Tweed. Transactions of the Botanical Society of Edinburgh 34: 348-360. Duncan, U. K. (1969). Isle of Tiree. 3rd—10th July. Proceedings of the Botanical Society of the British Isles 7: 636-637. Gopwin, H. (1975). The history of the British flora. 2nd ed. Cambridge University Press, Cambridge. HesLop Harrison, J. W., ed. (1941). A preliminary Flora of the Outer Hebrides. Proceedings of the University of Durham Philosophical Society 10: 228-273. HEsLop Harrison, J. W. (1949). Potamogetons in the Scottish Western Isles, with some remarks on the general natural history of the species. Transactions of the Botanical Society of Edinburgh 35: \—25. HeEsLop Harrison, J. W. & Crark, W. A. (1941). Hybrid Potamogetons on the Isle of Benbecula. Occasional ~ notes from the Department of Botany, King’s College, Newcastle upon Tyne, 2 2: \—4. HESLop Harrison, J. W. & Crark, W. A. (1942). A note on x Potamogeton suecicus Richt. Occasional notes from the Department of Botany, King’s College, Newcastle upon Tyne, 2 4: 4. HESLop Harrison, J. W. & HeEsLop Harrsion, J. (1950). A contribution to our knowledge of the flora of the Isles of Lewis, Harris, Killegray and Ensay. Transactions of the Botanical Society of Edinburgh 35: 132-156. HeEsLop Harrison, J. W., HESLop HARRISON, H., CLARK, W. A. & Cooke, R. B. (1941). The flora of the Isles of Coll, Tiree and Gunna (v.-c. 110b). Proceedings of the University of Durham Philosophical Society 10: 274-308. HEsLop Harrison, J. W., HeEsLop Harrison, H., CLARK, W. A. & Cooke, R. B. (1942). Further observations on the vascular plants of the Outer Hebrides (v.-c. 110). Proceedings of the University of Durham Philosophical Society 10: 358-367. HOLLINGSworTH, P. M., Preston, C. D. & GorNaLL, R. J. (1996a). Isozyme evidence for the parentage and multiple origins of Potamogeton Xx suecicus (P. pectinatus x P. filiformis, Potamogetonaceae). Plant systematics-and evolution 202: 219-232. HOLLINGSworTH, P. M., Preston, C. D. & GoRNALL, R. J. (1996b). Genetic variability in two hydrophilous species of Potamogeton, P. pectinatus and P. filiformis (Potamogetonaceae). Plant systematics and evolution 202: 233-254. Homes, N. T. H. & Wuitton, B. A. (1975a). Macrophytes of the River Tweed. Transactions of the Botanical Society of Edinburgh 42: 369-381. Homes, N. T. H. & Wuitton, B. A. (1975b). Submerged bryophytes and angiosperms of the River Tweed and its tributaries. Transactions of the Botanical Society of Edinburgh 42: 383-395. JOHNSTON, H. H. (1922). Additions to the Flora of Orkney, as recorded in Watson’s “Topographical Botany,” second edition (1883). Transactions of the Botanical Society of Edinburgh 28: 98-117. McCa.ium Wesster, M. (1978). Flora of Moray, Nairn & East Inverness. Aberdeen University Press. McNEILL, M. (1910). Colonsay. David Douglas, Edinburgh. PERRING, F. H. & RANDALL, R. E. (1972). An annotated Flora of the Monach Isles National Nature Reserve, Outer Hebrides. Transactions of the Botanical Society of Edinburgh 41: 431-444. PERRING, F. H. & SELL, P. D. eds. (1968). Critical supplement to the Atlas of the British Flora. Thomas Nelson & Sons, London. Preston, C. D. (1990). Potamogeton filiformis in Anglesey. Watsonia 18: 90-91. Preston, C. D. (1991). Potamogeton L., in PANKHUuRST, R. J. & Mutuin, J. M. (1991). Flora of the Outer Hebrides, pp. 129-133. Natural History Museum Publications, London. Preston, C. D. (1995). Pondweeds of Great Britain and Ireland. BSBI Handbook no. 8. Botanical Society of the British Isles, London. 342 C. D. PRESTON, P. M. HOLLINGSWORTH AND R. J. GORNALL Preston, C. D. (in press). Some overlooked specimens of Potamogeton x suecicus. B.S.B.I. Scottish Newsletter. Preston, C. D. & Crort, J. M. (1997). Aquatic plants in Britain and Ireland. Harley Books, Colchester. Preston, C. D. & Stewart, N. F. (1994). Irish Pondweeds V. Potamogeton X suecicus K. Richter in Co Donegal, new to Ireland. /rish Naturalists’ Journal 24: 485-489. Preston, C. D. & Stewart, N. F. (1995). Potamogeton X suecicus in Orkney. Orkney Field Club Bulletin 1995: 38-40. Preston, C. D., HOLLINGSWoRTH, P. M. & GorNALL, R. J. (1998). Potamogeton pectinatus L. X P. vaginatus Turcz. (P. x bottnicus Hagstr.), a newly identified hybrid in the British Isles. Watsonia 22: 69-82. RoYAL Botanic GARDEN, EDINBURGH (1983). Survey of aquatic vegetation on South Uist and Benbecula 25 July—5 August 1983. Unpublished report. Scott, W. & PAcMer, R. C. (1987). The flowering plants and ferns of the Shetland Islands. Shetland Times, Lerwick. TRAIL, J. W. H. (1901a). The flora of Buchan. Transactions of the Buchan Field Club 6: 69-162. TRAIL, J. W. H. (1901b). The flora of Buchan. Annals of Scottish natural history 1901: 164-176. Van Wuk, R. J. (1988). Ecological studies on Potamogeton pectinatus L. |. General characteristics, biomass production and life cycles under field conditions. Aquatic botany 31: 211-258. (Accepted February 1999) Watsonia 22: 343-352 (1999) 343 The past and present status of Moneses uniflora (L.) Gray (Pyrolaceae) in Scotland JZA, WRIGHT and P. S: LUSBY Royal Botanic Garden, Inverleith Row, Edinburgh, EH3 5LR ABSTRACT Moneses uniflora (L.) Gray (Pyrolaceae), One-flowered Wintergreen, is a nationally rare plant in Britain. It is now restricted to three vice-counties in Scotland. The history, decline and present status of the plant are described. Botanical collecting, deforestation and other land use changes are identified as causes of decline. Sensitive management and effective communication are vital for the survival of colonies in commercial forests. Keyworpbs: Woodland, conservation, forestry. INTRODUCTION Moneses uniflora (L.) Gray (Pyrolaceae), One-flowered Wintergreen (hereafter Moneses) is a nationally rare plant in Britain. It is not protected under Schedule 8 of the Wildlife and Countryside Act (1981) but is classified as vulnerable in the British red data book for vascular plants (Perring & Farrell 1983). The species has been previously reported from at least 15 vice-counties with doubtful records from a further threc, including the English vice-county of Westmorland (69) (not included in Fig. 1 which shows the distribution of the species). There has been a marked decline and it is presently recorded in twelve 10-km squares in three vice-counties (Table 1). TABLE 1. PRESENT SITES AND POPULATION SIZES OF MONESES UNIFLORA IN SCOTLAND Site Vice-county 10-km square Population sizes Culbin Forest 95 NH/9.5 A Culbin Forest 95 NH/9.6 [No Jah (CAE Culbin Forest 95 NJ/0.6 A Old Grantown Wood 95 NJ/O.2 A Burgie Wood 95 NJ/L.S B Lethenhill 95 NJ/1.5 A Glen Affric 96 NH/2.2 A,C Strath farrar 96 NH/2.3 A Strathfarrar 96 NH/3.3 A Glen Einich 96 NH/9.0 A Rothiemurchus 96 NH/9.0 A Loch Morlich 96 NH/9.0 A Loch Loy 96 NH/9.5 A Abernethy Forest 96 NJ/0.1 A The Mound 107 NH/7.9 A Balblair Wood 107 NH/8.9 D Vice-counties: 95 Moray, 96 Easterness, 107 Sutherland East. Population sizes: A -— 1-199 rosettes, B — 200—499 rosettes, C — 500-999 rosettes, D — >1000 rosettes. 344 J. A. WRIGHT AND P. S. LUSBY FiGuRE 1. The distribution of Moneses uniflora in the British Isles. Symbols: @ — 1985 onwards, O — 1970-1984, X — pre-1970 Moneses is a member of the Circumpolar Boreal-montane element of the British flora and its occurrence in Britain represents the westernmost limit of its distribution in Europe (Preston & Hill 1997). The species ranges from central and northern Europe, south to the southern Adniatic and east to the Novosibirsk region in Russia. It also occurs in abundance in the eastern United States of America. In central and northern Europe Moneses mainly grows in humid spruce woods but also occurs in drier forest dominated by Scots Pine (Pinus sylvestris). The field layer of these woods is dominated by dwarf ericaceous shrubs. In the Swiss Alps Moneses is a characteristic species of subalpine spruce woods with Lycopodium annotinum, Listera cordata and Linnaea borealis (Ellenberg MONESES UNIFLORA IN SCOTLAND 345 TABLE 2. VEGETATION OF MONESES UNIFLORA SITES IN SCOTLAND Site NVC Community (where identified) or other vegetation type Lethenhill MI5b* Balbair Wood W18a* Little Ferry W18s The Mound W18 Burgie Wood W18a* Culbin Forest W18a* Culbin Forest Corsican Pine (Pinus nigra var. maritima)* Glen Emich WI18b Strathfarrar (both sites) W18b Rothiemurchus WI18b Old Grantown Wood W18c* Abernethy W18d Glen Affric (both sites) W18d * = forest plantation. Key to NVC communities (see Rodwell 1991a,b) M15 Scirpus cespitosus — Erica tetralix wet heath; b = typical subcommunity. W18 Pinus sylvestris — Hylocomium splendens woodland subcommunity, a = Erica cinerea — Goodyera repens subcommunity; b = Vaccinium myrtillus — V. vitis-idaea subcommunity; c = Luzula pilosa sub- community; d = Sphagnum capillifolium — S. quinquefarium subcommunity. 1988). On the German island of Riigen, Moneses occurs in moss-rich pinewoods on young wooded dunes (Meusel 1951), habitats similar to the Scottish coastal sites in Moray and East Sutherland. However, the plant communities of most of the Scottish sites (Table 2), are similar, both floristically and structurally, to the pine/birch forests of western Norway (Aune 1977). Across its total range Moneses occurs more abundantly in humid spruce forests rather than in drier pinewoods. At one locality near Elgin, Moneses has been found in a wet-heath (NVC MI5b) (Table 2), but many plants did not appear healthy. There are no fossil records of Moneses or any other Pyrolaceae in Britain or Ireland. This paper reports the results of a field and literature survey carried out by the Scottish Rare Plant Project of Scottish populations of Moneses. Original sources were consulted to determine as precisely as possible the location of old records. Vice-county recorders, nature reserve wardens, staff of various conservation organisations, foresters, local botanists and others have been consulted in gaining up-to-date information on the status of different populations. Records were obtained from 24 herbaria. Unpublished and manuscript sources have been checked for additional records. DISCOVERY OF MONESES UNIFLORA IN BRITAIN Moneses was known only as a cultivated garden plant to herbalists in Britain in the sixteenth and seventeenth centuries (Gerard 1597; Parkinson 1640). Miller (1763) reported that Moneses grew “naturally in shady woods in the northern parts of Europe’, but he did not know it was native in Britain. Philip Miller cultivated Moneses in 1748, making this the earliest record of the plant growing in Britain, albeit as an introduction (Murray 1799). The first native record is of doubtful authenticity. There are two specimens in the herbarium of Sir J. E. Smith (LINN) labelled “From the western Isles of Harris and Bernera gathered in 1783 by Jas. Hoggan’. These were sent to Smith in 1793 by R. Gotobed. There are no other records of Moneses from the Western Isles (Pankhurst & Mullin 1991). James Brodie found Moneses in pinewoods near Brodie House, Forres, in 1792 (Clarke 1900); undated specimens collected by Brodie survive in E. The same year James Hoy, secretary and librarian to the Duke of Gordon at Gordon Castle, Fochabers, sent a specimen collected locally to the Linnean Society (Sowerby 1794). 346 J. A. WRIGHT AND P. S. LUSBY SCOTTISH SITES OF MONESES UNIFLORA WEST SUTHERLAND (V.C. 108) A single collection by Collins from Scourie in 1905 (LRS) constitutes the northernmost record for Moneses in Britain (Fig. 1). EAST SUTHERLAND (V.C. 107) The largest and best known population in Bnitain is at Balblair Wood near Golspie; several thousand plants occur in a Scots Pine plantation covering about 40 ha. The first record was by Crawford in 1890 (Kenworthy 1976). Specimens labelled ““Golspie” and “near Golspie” may be from this locality, but those labelled “Ferry Wood” and “Little Ferry” probably came from Ferry Wood which is separated from Balblair by a road. Certainly Anthony (E) made this distinction. The present plantation of Scots Pine at Balblair is just over 75 years old and was established on the site of native pinewood which suffered windblow. The wood is owned by Sutherland Estates. Its conservation importance was recognised in 1970 when, in conjunction with Sutherland Estates, the Scottish Wildlife Trust established a wildlife reserve. Balblair Wood is also part of the Loch Fleet Site of Special Scientific Interest notified by The Nature Conservancy Council in 1975. A Nature Research Agreement has recently been secured by Scottish Natural Heritage with the intention to declare the area as a National Nature Reserve (F. Symonds, pers. comm., 1998). Moneses was also collected from The Mound, 2-5 km north-west of Balblair, where it was first recorded by Foggit (Kenworthy 1976) and Crawford (E) in 1900. Druce noted that Moneses grew “in immense quantity in a larch wood near Golspie (a second locality)” (OXF). Druce’s 1923 collection was the last known record from The Mound until Symonds rediscovered a small patch on the south side of The Mound within the Mound Alder Woods National Nature Reserve in 1997. There are several collections (ABD, BM, E, RNG) dated 1939 from Cambusmore on the western shore of Loch Fleet. These are the only known records from this locality. There is also a single record from Dunrobin, north of Golspie (Watson 1837). EAST ROSS-SHIRE (V.C. 106) There are no extant sites in East Ross-shire. Records from the chief locality, variously known as “Knock Farnl’, “Coul Wood” or “The Cat’s Back’, near Strathpeffer, span the period 1830 to 1872. The most precise details are on a specimen dated 1835 (ABD) “In the Coul fir wood, about a mile {2 km] to the west of the Strathpeffer pump-room, ... in two or three large patches”. Gordon (1867, ms letter to Dr J. Mitchinson, ELN) reported that “the firwood has been cut down, so I suppose the plants have perished for lack of shade’. Hillhouse (1889) declared Moneses extinct at Knock Farril. However, Moneses was found at Strathpeffer in 1966 by Duncan (1980), but a programme of tree felling and replanting began there in 1968.-Selby monitored the status of Moneses and made his last sighting about 1970. Brebner & Hulme (pers. comm., 1994) knew a few plants beside a track near Strathpeffer Youth Hostel durng the mid-1970s, but these were unwittingly destroyed by widening and re-routing of the track. A specimen collected by McRae dated 1905 (E) and labelled from the Black Isle may represent a second locality but is as likely a vague reference to Strathpeffer which is close to the Black Isle. The plant has never been refound at Torr Achilty, south-west of Contin, where it was recorded in 1863 (Duncan 1980). MID EBUDES (V.C. 103) Moneses was listed from Torosay, east Mull, “according to Mr Middleton at Achnacroish” by Clerk (1845), but has not been included in any other plant lists for Mull. MAIN ARGYLL (V.C. 98) Marshall reported Moneses from Kilmory Estate near Lochgilphead, Argyll: “Sir John Campbell- Orde, Bart recently showed me this plant growing on his estate near Lochgilphead, and assured me it was not an introduction. This appears to be the first certain station for it in west Scotland” (Marshall 1896). No other record is known. EASTERNESS (V.C. 96) Moneses was collected by Ballie in 1890 (BM) “in a fir wood in rather boggy ground”. In 1990 it was recorded from the margin of Loch Loy by North. Two small colonies, separated by a few MONESES UNIFLORA IN SCOTLAND 347 metres, are extant. The habitat is an open mixed Scots Pine and Downy Birch (Betula pubescens) wood. There are extant populations in Strathfarrar and Glen Affric. Moneses was found in Strathfarrar by Miss Fraser Lovat in 1867 (Farquharson & Selkirk 1868). There was no other record from Strathfarrar until 1980 when Cameron discovered one of the extant colonies, cast of Loch Beannacharan, at about 275 m altitude, in native pinewood. The flora is dominated by Vaccinium myrtillus and observations over the last ten years suggest an increase in density of this species, while the number of Moneses rosettes has declined. It may be necessary to reduce the competition from V. myrtillus to preserve this colony. The second population, on a trackside towards the eastern end of Strathfarrar, was discovered in 1992 by Mrs E. Lennard. The track is used at present for access to a hydroelectric installation, so this colony is also vulnerable. In Glen Affric, Moneses was found in 1975 by Crawiey (E), north-east of Loch Beinn a’ Mheadhoin; this record is not included in McCallum Webster (1978). Despite recent searches, this colony has not been refound and may have been eliminated by tall heather (M. Barron, pers. comm., 1995). In 1988 and 1990 two populations were found on the margins of lochans on the south side of Glen Affric. Here the plants are growing just above summer water levels and are occasionally submerged in winter. One population comprises four small colonies distributed over some 200 m; the other is a single colony of a few rosettes. A further population in Glen Affric was found by Lennard on the lower slopes of Creag Dubh above Cougie (Lennard, pers. comm., 1993). A colony from a “field near the dam at Kingsmill, Inverness” found by Galloway before 1888 is no doubt lost and another from Strathdearn (McCallum Webster 1978) has not been recorded since. Records from Abernethy are often imprecise but Traill (1910) and Davidson (ABC) recorded Moneses from Loch Mallachie on the southern margin of Abernethy forest. No recent records from this area are known. An extant population within Abernethy Forest was found by Horn in 1988 and consists of fewer than 100 rosettes scattered over an area of about 25 m’. The vegetation is dominated by Calluna vulgaris and Vaccinium myrtillus with frequent Carex nigra and C. panicea in wet channels. A specimen labelled “Boat of Garten” (1919, (E)) may be from Abernethy Forest. The forests of Rothiemurchus and Loch an Eilean have been known sites for Moneses since 1882 (Keith & Groves (FRS, BM)). MacMillan (1907) reported the plant “in some abundance in the woods at the south-west end of the Loch’. Hillhouse (1889) noted that it was “disappearing from Rothiemurchen [sic] ... from the rapacity of collectors”. Over 50 plants were gathered between 1882 and 1894, but at this site, at least, Moneses was not collected to extinction. One small colony is extant in Rothiemurchus, in Glen Einich. About 20 rosettes occur at the base of a dead pine tree in atypically open, dry conditions. The associated vegetation is dominated by Vaccinium myrtillus, with V. vitis-idaea and common pleurocarpous mosses. The population occurs between two drainage channels which direct water away from the plants. This could account for lack of vigour in this colony. Records from Loch Morlich and Glen More are first represented by a specimen collected by King and exhibited to the Natural History Society of Glasgow in 1885 by Boyd (Strton 1887). Since then populations have been recorded intermittently. Two small colonies have been washed away by flash floods and another destroyed by Sitka Spruce plantation (D. Ross, pers. comm., 1990). In 1994 a population was found near Loch Morlich by Jones. A colony at Kincraig, between Aviemore and Kingussic (McCallum Webster 1978) and onc in Glen Feshie (Steven & Carlisle 1959), have not been confirmed by our survey. MORAY (V.C. 95) Moray is considered the headquarters for Moneses in Scotland. Although the extent of all colonics does not approach the size of the Balblair population, more populations have been recorded in Moray than in any other vice-county. After Brodie’s discovery in 1792 there are no further records from Brodie House other than a specimen in DBN dated 1798 (Nelson 1995). Several botanists reported the plant’s disappearance at this site which was attributed to tree felling and gorse invasion (Bishop 1826; Brichan 1842; Ogilvie 1845). Therefore the original site was lost within about 30 years of its discovery. The second record from Moray is in Gordon’s Collectanea for a Flora of Moray (1839): “Discovered by John Lawson esq. about 20 years ago [1819] in the oakwood, near Aldroughty. It was afterwards lost sight of until 1836, when a few specimens were gathered by J. Shier, esq. and 348 J. A. WRIGHT AND P. S. LUSBY pupils’. One specimen is in ELN. This locality was recorded by Hooker (1830) as “Knock of Alva [Alves]. Other localities near Elgin are Loch Avain (Todd 1867, 1887 (CGE)) and Roseisle Forest, near Burghead, recorded by Cuthbertson in 1974 (McCallum Webster 1978). In 1908 the plant was found in Balnacoul Wood, west of Fochabers by Watson (Burgess 1935) but no specimen has been traced. This site is now a forestry plantation. There are five extant localities for Moneses in Moray. Chief of these is the afforested sand dune system of Culbin Forest which occupies nearly 3000 ha of the southern coast of the Moray Firth between Nairn and Forres. Planting began on Culbin Sands in 1839 but was most extensive between 1922 and 1963 (McCallum Webster 1968). Without cxact details it is not possible to ascertain whether records were from within Culbin Forest or whether they were from woodland outside its present boundary. Specimens labelled ‘“‘Forres” dating from 1840 to 1871 could be from woodland existing prior to afforestation, but an 1869 collection from “Clunie Hill” (south-east of Forres) (Brown (ABD, BM)) 1s clearly a separate locality. No subsequent records of the latter site are known. Burgess discovered three stations in Dyke (0-8 km north of Brodic) in 1901 which may be within Culbin Forest and a specimen from “‘Snab Wood” collected by Patton in 1923 (GL) could also be from Culbin. Miss Mary McCallum Webster knew Culbin Forest and the localities of Moneses better than any other botanist but was always vague regarding sites. Precise details were never handed on in her lifetime. Those fortunate enough to be shown a colony were led on a deliberately tortuous route along forest tracks which made relocation practically impossible! Consequently, it is not certain which sites discovered since the death of Miss McCallum Webster coincide with the ones she knew. To date six colonies have been found in Culbin Forest by North, Farrell, White, Edelsten and Young. Population sizes range from less than 20 rosettes to several hundred, and habitats vary from very wet to dry (Table 1). : Records from Burgie date from 1870 with collections by Innes and Keith (ABD, BM, FRS, K, OXF). It was found again in 1910 by MacGregor (Burgess 1935) and in 1920 by McCallum Webster (E). A small colony was found under Scots Pine and Larch (Larix decidua) at the edge of Burgie Wood (now a large forest plantation) by Lusby in 1993. Immediately to the east of Burgie Wood at Lethenhill, Matthews found a small colony of Moneses among scattered Scots Pine and Juniper (Juniperus communis) in wet heath (Tables 1 & 2). This was the wettest site for Moneses; a number of rosettes displayed veinal chlorosis. Records from Grantown-on-Spey (MacKechnie 1954, (E)) and Castle Grant (McCallum Webster 1978) probably refer to Old Grantown Wood where the plant is extant but in a precarious condition. Timber extraction has destroyed some colonies and grazing threatens others. Moneses is accompanied by abundant Linnaea borealis, Goodyera repens and Ptilium crista-castrensis at this site. In the early 1950s Lennard recorded Moneses from “open heathy moorland but close to natural pinewood” (Lennard, pers. comm., 1993) about 3 km east of Nethy Bridge and south of Craigmore Wood but this has not been confirmed by our survey. BANFFSHIRE (V.C. 94) Moneses was collected between Dufftown and Drumuir in 1840 and 1890 (Anon (E)). Blizzard Bell (Greville 1841) and Dickie (1860) recorded the plant from Mortlach which could also be the same locality. It was found in Glen Livet by Keith in 1870 (CMM). NORTH ABERDEEN (V.C. 93) The earliest record for v.c. 93 is a collection from Haddo by Stewart in 1893 (E), but it has not been recorded subsequently from this site. Farther west, Moneses was collected in 1861 (Anon (ABD)) from a wood between Rhynie and Clatt. Since then, any habitat suitable for Moneses has been lost to agriculture. Pirie (1906) recorded a small patch of Moneses from Bin Wood, 3 km north-west of Huntly, but recent searches have failed to refind the plant. SOUTH ABERDEEN (V.C. 92) _Moneses has been recorded from possibly three localities. However, the collection from Braemar by White in 1877 (ABD) and the record from Ballochbuie Forest by Steven & Carlisle (1959) could MONESES UNIFLORA IN SCOTLAND 349 be the same locality. An undated specimen from “Burnwood, Kincardinshire” (Anon (ABD)), is probably from Burn of Wood near Kirkton of Glenbuchat. ANGUS (V.C. 90) Don recorded Moneses as “rare” in the Clova Mountains (Don 1813) but Gardiner (1848) quesuioned this locality. Don offered Moneses for sale from his Forfar nursery in 1813 but whether the source of Don’s cultivated material was the Clova Mountains can only be guessed. This is the only known Angus record apart from a dubious specimen in OXF. This record is not included in Biss 1: EAST PERTH (V.C. 89) Moneses was first reported from “near Perth ... in considerable abundance” by Bishop (1826) but less than a century later Barclay (1908) blamed collectors for the “greatly lessened quantity”’. The plant was avidly collected from Scone between 1825 and 1833 (e.g. Gardiner 1848). Drummond-Hay holds the record for the largest number of specimens (25) on a single sheet. From this period 69 herbarium sheets and another 20 collections have been traced. Sim (1859a) stated that Moneses “was found under the trees, among moss and grass, sparingly distributed over an area of about two acres [1 ha] ...” but revisited the site the next summer and “could only obtain a few rather stunted specimens” (Sim 1859b). However, the subsequent year he was more successful, advertising, “specimens of Moneses grandiflora ... will be supplied” (Sim 1860). Several botanists including Hillhouse (1889), White (1898) and Barclay (1908) lamented the decline of Moneses at Scone and the last record traced from Scone is a specimen dated 1922 (PTH). Records from Muirward Wood and New Scone are probably the same site and represent a second locality. Specimens were collected by Sadler in 1857 (E) (Balfour 1902) and the last record is Drummond-Hay’s specimen of 1869 (PTH). MID PERTH (V.C. 88) Specimens collected from Methven by McNab in 1836 and Campbell in 1837 are in CGE and E respectively. However, Sadler and others failed to locate the plant there on an excursion in 1857 (Balfour 1902). A recent record for Craigvinean Forest in Strathtay requires confirmation (Turl, pers. comm, 1998). MIDLOTHIAN (V.C. 83) Learmonth (1841) recorded Moneses from “Harburn Firwood”, just south of West Calder. The rarer plants he listed have disappeared from this wood. KIRKCUDBRIGHTSHIRE (V.C. 73) This is the most southerly Scottish record. It was reported by Hillhouse (1889) as “Extirpated from Woodhead Hill, Traqueer [sic], Dumfnesshire’. Woodhead Hill is within Mabie forest which is in the parish of Troqueer, south-west of Dumfries. No records of Moneses can be traced in the Floras of Dumfnes or Kirkcudbnghtshire. WESTMORLAND (V.C. 69) Borrer recorded Moneses from Bardsea, near Ulverston, where “he sought it unsuccessfully, not withstanding ‘a very particular direction’ by Wright of Keswick” (Watson 1849). Baker (1885) doubted this record. The only English specimen is labelled ‘““Westmorland” (det. Shillito 1820 (LIV)). DISCUSSION Moneses has been lost from at least 28 Scottish sites. More than half of these losses have been from north-eastern Scotland. There has apparently been not only a decrease in the abundance of the Species but also a considerable contraction of its British range. As Moneses is easily overlooked, and appears to exist mostly in very small populations, it is possible that the plant is under-recorded, especially in the larger remnants of native pinewood. However, discoveries distant from well known areas for the plant are rare. 350 J. A. WRIGHT AND P. S. LUSBY TABLE 3. HERBARIUM COLLECTIONS* OF MONESES UNIFLORA FROM SCOTLAND (FROM 24 BRITISH HERBARIA) East Easter Decade Perthshire Moray Sutherland Ross Speyside Grampian Other Total 1790s 3 3 1820s ] 1 1830s 3] ] 3] 63 1840s 30 l 3 l l 36 1850s 9 9 1860s 14 2} l l 18 1870s 3 26 ] 2 2 34 1880s ] D 4 i 1890s - 9 ] 18 2 34 subtotal 1790-1899 205 1900s ] 9 ] 1] 1910s 2 2 1920s ] 2 6 9 1930s 1 9 10 1940s l l 2 1950s 2 3 5) 1960s ] 5 6 1970s ] ] ] 3 1980s 2. « l 3 subtotal 1900-1989 Sil Undated 8 Dp 2 ] 2} ] 0 16 Total 98 52 42 37 29 3) 9 2q2 *A collection is defined as a plant or plants gathered from a named locality by one collector (where recorded) attached to a single herbarium sheet. Where two or more sheets with the same date, locality and collector have been distributed to different herbaria, these have been counted as separate collections. Recently discovered populations have been spotted by chance, not by systematic searches of “lost” sites. Moneses is apparently able to persist for a considerable length of time in small, fragmentary populations, for example in Burgie Wood and The Mound. The small size of nearly all extant populations (Table 1) renders the plant vulnerable to habitat disturbance, a fact borne out by reports of Moneses disappearing from sites due to tree-felling or track-widening. Secrecy from those who should know about localities, rather than ensuring the plant’s protection, is often more likely to result in its unwitting destruction, especially in managed plantations. Although botanical collecting during the mid- to late-1800s (Table 3) was a major cause of decline of Moneses in Britain, this threat has now subsided. With the exception of traits that render Moneses sensitive to changes in its habitats, the main current threat 1s poor communication between conservationists, botanists and land managers. Forestry practices have had both positive and negative effects on Moneses populations. Where modern plantations of exotic conifers have replaced old pinewoods, Moneses has been drastically reduced or lost, whilst some old Scots Pine plantations have either been colonised by Moneses or the plant has been introduced with stock. The latter may have been the origin of some populations on estates with large conifer plantings, for example Scone and Kilmory Estate, Lochgilphead. Most of the British population of Moneses occurs on land which is managed for timber production. Therefore the future of the plant depends on sensitive management within areas of forests. In furthering this aim, considerable progress has been made in recent years by Scottish Natural Heritage (North Area) and Forest Enterprise maintaining regular and effective communication regarding the whereabouts of Moneses populations. MONESES UNIFLORA IN SCOTLAND 351 ACKNOWLEDGMENTS We wish to thank Shaun Ince, Jonathan Hughes, Jonathan Stacey, Ben Averis, Margaret Lennard, John Edelsten, Fraser Symonds and others who have helped us find Moneses at certain sites. Thanks also to Gordon Rothero for producing the distnbution map using DMap. We also thank Douglas McKean, Tim Rich and the library staff of the Royal Botanic Garden, Edinburgh. REFERENCES Aung, E. I. (1977). Scandinavian pine forests and their relationship to the Scottish pinewoods, in Bunce, R. G. H. & Jerrers, J. N. R. eds. Native pinewoods of Scotland. Institute of Terrestrial Ecology, Cambridge. Baker, |. G. (1885). Flora of the English Lake District. George Bell & Sons, London. Ba.rour, J. H. (1902). Botanical excursions made by John Hutton Balfour. Notes from the Royal Botanic Garden Edinburgh 8: 21-481. Barciay, W. (1908). Address to Society. Proceedings of the Perth Society of Natural Sciences 4: 189-194. BisHop, D. (1826). Rare native plants found in Perthshire. Edinburgh philosophical journal 14: 180-181. BrRIcHAN, J. B. (1842). Notes on the British Pyrolae. Phytologist 1: 26. Burcess, J. J. ed. (1935). Flora of Moray. Moray Field Club, Elgin. CiarKE, W. A. (1900). First records of British flowering plants. West, Newman & Co., London. Cuierk, D. (1845). The Parish of Torosay. Statistical account of Scotland: Renfrew and Argyle 7: 277-296. DickiE, G. (1860). Botanist’s guide to the Counties of Aberdeen, Banff and Kincardineshire. A. Brown & Co., Aberdeen. Don, G. (1813). Account of the native plants in the County of Forfar, and the animals to be found there, in HEApRICK, J. General view of the agriculture of the County of Angus or Forfarshire. Neill & Co., Edinburgh. Duncan, U. K. (1980). Flora of East Ross-shire. Botanical Society of Edinburgh, Edinburgh. ELLENBERG, H. (1988). Vegetation ecology of Central Europe. Cambridge University Press, Cambridge. FARQUHARSON, J. & SeLKiRK, A. M. (1868). Notes of a visit to Strath Glass and its tributary glens. Transactions of the Botanical Society of Edinburgh 9: 474-479. GARDINER, W. (1848). Flora of Fofarshire. Longman, Brown, Green & Longmans, London. GERARD, J. (1597). The Herball or Generall historie of plants. John Norton, London. Gopwin, H. (1975). The history of the British flora, 2nd ed. Cambridge University Press, Cambridge. Gorpon, G. (1839). Collectanea for a Flora of Moray. G. Gordin, Elgin. GrevILLe, R. K. (1841). Communications. 4th and Sth Annual Report of the Proceedings of the Botanical Society of Edinburgh. Edinburgh. HILLHouse, W. (1889). The disappearance of British plants. Journal of botany 27: 359-365. Hooker, W. J. (1830). Flora Scotica. Archibald Constable & Co. and Hurst Robinson & Co., London. LEARMONTH, W. (1841). Parish of West Calder. Statistical account of Scotland 1: 304-309. Kenwortny, J. B. (1976). John Anthony's Flora of Sutherland. Botanical Society of Edinburgh, Edinburgh. McCatitum Wesster, M. (1968). A checklist of the flora of Culbin State Forest. M. McCallum Webster, Moray. McCa.ium Wesster, M. (1978). Flora of Moray, Nairn and East Inverness. Aberdeen University Press, Aberdeen. : MAcMILLAN, H. (1907). Rothiemurchus. Dent & Co., London. MarsuHa lt, E. S. (1896). Moneses grandiflora in Argyll. Journal of botany 34: 400. MeuseEL, H. (1951). Vegetationskundliche Studien tiber mitteleuropaische Waldgesellschaften. Berichte Deutschen Botanischen Gesellschaft 64: 222-240. MILLER, P. (1763). Gardener’s dictionary, 5th ed. John Rivington, London. Murray, C. (1799). The British garden: a description of hardy plants, indigenous, or cultivated in the climate of Great Britain. S. Hazard, Bath. NELson, E. C. (1995). Scottish botanical history preserved in the National Botanic Gardens, Glasnevin, Dublin. Scottish naturalist 107:137-162. OaiLvie, W. (1845). The Parish of Dyke and Moy. Statistical account of Scotland; Banff Elgin, Nairn 13: 218. PANKHURST, J. & MULLIN, J. M. (1991). Flora of the Outer Hebrides. Natural History Museum, London. PARKINSON, J. (1640). Theatrum Botanicum. Thomas Cotes, London. PERRING, F. H. & FARRELL, L. (1983). British red data books 1: vascular plants, 2nd ed. Royal Society for Nature Conservation, Lincoln. Pirie, J. (1906). The Parish of Cairney. J. Pirie, Banff. PRESTON, C. D. & Hitt, M. O. (1997). The geographical relationships of British and Irish vascular plants. Botanical journal of the Linnean Society 124: 1-120. 352, J. A. WRIGHT AND P. S. LUSBY RopweLL, J. S., ed. (1991a). British plant communities. Vol. 1. Woodlands and scrub. Cambridge University Press, Cambridge. Ropwe Lt, J. S., ed. (1991b). British plant communities. Vol. 2. Mires and heaths. Cambridge University Press, Cambridge. Sim, J. (1859a). Plants of Perth. Phytologist 3: 33-44. Siu, J. (1859b). Botanical rambles in June 1859. Phytologist 3: 359-362. Sim, J. (1860). Botanical notes, notices, and queries. Phytologist 4: 317. Sowersy, J. E. (1794). English botany, Vol. 2. J. E. Sowerby, London. STEVEN, H. M. & Caruisve, A. (1959). The native pinewoods of Scotland. Oliver & Boyd, Edinburgh. STIRTON, J. (1887). Communication, summer session, 1885. Proceedings of the Natural History Society of Glasgow, New Series 1: Ixv. TRAILL, J. W. H. (1910). Botanical notes. Annals of Scottish natural history 76: 253. Watson, H. C. (1837). New Botanists’ Guide. Longman, Brown, Green & Longmans, London. Watson, H. C. (1849). Cybele Britannica. Longman & Co., London. White, F. B. W. (1898). The Flora of Perthshire. Perthshire Society of Natural Science, Edinburgh. (Accepted February 1998) Watsonia 22: 353-364 (1999) 353 Changes in the distribution and abundance of Himantoglossum hircinum (L.) Sprengel (Orchidaceae) over the last 100 years PAD ECARE YE Institute of Terrestrial Ecology, Monks Wood, Abbots Ripton, Huntingdon, Cambs, PE17 2LS ABSTRACT 1. Species are likely to be most sensitive to climate change at the geographic limits of their distribution. The behaviour of such populations may therefore be a predictor of the response of the species to global change. 2. The northern limit of the Lizard Orchid, Himantoglossum hircinum, occurs in England, where 16 populations are currently known. British records over the past 100 years are particularly accurate for this species and it was one of the first for which changes in distribution were linked to an amelioration of climate (Good 1936). 3. Early records were checked to confirm the rise in population number in the early part of the century. Analysis of more recent data showed that this was followed by a sharp decline and numbers have only been rising again over the last decade. The range expanded with the earlier increase in population number, but did not contract as populations were lost. 4. Data collected between 1977 and 1998 in the largest population allowed flowering probability and seed production to be correlated with rainfall during the growing season. Analysis of the resulting model showed that both observed rises in population number followed periods during which the seasons for vegetative growth had been wet. 5. Populations have become both larger and more persistent due to an increased interest in conservation. 6. Changes in the abundance of H. hircinum are likely to depend on other factors, including patterns of human activity, as well as on climate change. KeEyworbs: rainfall, life-cycle, range-limits, seed production, climate change. INTRODUCTION There is currently a great deal of research being undertaken to assess the effects of climate change on the distribution and abundance of individual species (e.g. Carey & Brown 1994; Sykes et al. 1996; Parmesan 1996). This research is based on the theory that ultimately the distribution of a species is limited by its physiological responses to climate. One of the earliest exponents of this theory was Ronald Good who in 1931 proposed his “Theory of Tolerance”. He supported his theory with a paper on the distribution of the Lizard Orchid Himantoglossum hircinum (L.) Spreng. (Good 1936). In this paper he suggested that the increase in the distribution, and also the numbers of H. hircinum records, in the early part of this century were due to a climate change. Specifically, Good indicated that it had been “‘an amelioration of winter and spring temperatures and a slight increase in the preponderance of winter rain” that had made England similar to western France, which 1s at the heart of the species’ range. If Good was correct, we might expect that the distribution and abundance of H. hircinum in England would have increased further if the climate continued to be amenable. Furthermore, at least one climate scenario (Viner & Hulme 1994) suggests that England will become more like south-western France in the coming decades. Therefore, if H. hircinum does respond as Good suggested, we might expect the distribution and abundance of the species to increase (Carey & Brown 1994). In this paper Good’s theory is re-evaluated by re-analysing the early records of H. hircinum and also by adding records from the years 1934-1998. *Address for correspondence: E-mail: P.Carey @ite.ac.uk 354 P. D. CAREY DATA AND METHODS HISTORICAL DATA The records used in this paper come from many sources. The Biological Records Centre (B.R.C.) at Monks Wood holds a historical database for species in Great Britain and has provided the majority of data. Within the B.R.C. file on Himantoglossum hircinum (L.) Sprengel is a large body of information collated by Mrs G. Crompton and L. Farrell to whom I am indebted. This information includes copies of correspondence between Good and various botanists within England and abroad as well as the notes he used to produce the 1936 paper. The record cards from the B.R.C. hold information on the date the species was seen, where it was seen and by whom. There are also sections for information on habitat and notes. The notes section often contains information on the number of flowers or plants seen. Interpreting the records was not straightforward. It was not uncommon to have records from the same site in the same year given by different individuals who use different names for the location as well as different map references. Despite these inconsistencies it is almost always possible to determine which records are the same. I am fortunate that most of this work was done by others before me, by Good for the earlier records, and by G. Crompton and R. Fitzgerald for all records. I have added to their records by carefully re-reading the record cards, checking sources and correspondence on the species. The most modern records were supplied by M. Wigginton at the Joint Nature Conservation Committee and from personal observations. The records suggest that two populations have arisen from garden escapes but there may be others. These two are included here. Several populations are due to “well-meaning naturalists’’ who transplanted whole plants to new sites and counties. The number of these “introductions” is impossible to estimate but any records that are known to have been due to introductions have been discounted from this analysis. CONTINENTAL DISTRIBUTION Himantoglossum hircinum is spread over much of southern Europe and a map of this distribution is published (Meusel et al. 1965) and also digitised on to an Atlas Flora Europaea grid (Carey et al. 1993). The species is especially common in the wine-growing regions of France where it is noted as a roadside weed growing on a wide range of neutral to basic substrates including broken concrete. DEMOGRAPHIC CHARACTERISTICS “How long do Himantoglossum hircinum seeds take to germinate?” and “how long does the protocorm which develops remain underground before a leaf is produced?” are questions that, as yet, are not fully resolved. A recent study (P. Carey & H. Scott unpublished data) following the methodology of Rasmussen and Whigham (1993), has shown that 0-10% of seeds had signs of mycorrhizal infection (the first stages of germination) in the first autumn and winter after they were produced in 1996 whereas 10-40% of seeds produced in 1997 became infected almost immediately. The difference between the two autumns was that 1996 was dry and 1997 was wet. At the time of wniting no seedlings have been noted from either 1996 or 1997 seeds. There is anecdotal evidence to suggest that seedlings appeared three years after an isolated plant flowered for the only time (Good 1936). A matrix of plant states (Table 1), calculated from data collected from two permanent plots at the largest population in England from 1985-1994 (P. Carey & N. Stewart in prep.) indicates the proportion (0-59) of plants remaining at this “seedling” stage from year to year is very high. The low death rate of large and medium plants in the transition matrix (Table 1) also indicates that H. hircinum can be a long-lived plant. Some individuals first censused in 1987 are still alive and flowering today (L. Farrell, pers. comm., N. Stewart, pers. comm.). Many of these plants flower in successive years but, more importantly for this study, many do not. One individual plant in Sussex flowered in 1984, then again in 1995 and 1996, but never in between. It is also not unusual for plants to remain underground as tubers during any one year (N. Stewart, pers. comm., L. Farrell, pers. comm. and pers. obs.). In this paper the growing season is defined as September to August. The plant is typically wintergreen, emerging with the autumn rains anywhere from late August to April (G. Crompton, pers. comm, L. Farrell, pers. comm.; N. Stewart, pers. comm. & pers. obs.). Most plants are DISTRIBUTION AND ABUNDANCE OF HIMANTOGLOSSUM HIRCINUM 355 TABLE 1: TRANSITION MATRIX FOR MOVING FROM ONE LIFE STAGE TO ANOTHER FOR HIMANTOGLOSSUM HIRCINUM. TRANSITIONS ARE TAKEN AS THE MEAN TRANSITION PROBABILITIES FROM NINE YEARS OF DATA FROM TWO PERMANENT 10 x 10 M PLOTS AT THE LARGEST POPULATION IN ENGLAND. year t+] Plant size seedling small medium large death year t seedling 0-59 0-18 0-01 0-00 0-22 small 0-15 0-50 0-28 0-01 0-06 medium 0-03 0-10 0-47 0-38 0-02 large 0-01 0-07 0-18 0-71 0-03 apparent by November and only very small plants (seedling category in Table 1) emerge after this. The plants flower from late June to late July. Data collected from three permanent plots within the largest population in England between 1977 and 1994 (Carey & Stewart in prep.) were compared with climate data from the nearest weather station at Manston which is approximately 8 km north of the population. These data showed that the probability of a large plant flowering (Y) in growing season t is related (R’ = 61-8, F = 25-29, p<0-001) to the precipitation (X) in the months September—Apnil (the vegetative growth phase) in growing season t-1 (Equation 1). Therefore the flowering “initial” is likely to be determined by the size of the tuber set in the growing season before the growing season 1n which flowering takes place. Y, = 0-382 +0-0019X, , Equation 1 The production of seed pods was noted in the years 1988, 1989, 1991, 1992 and 1997. The number of pods produced per flowering spike (S) in growing season t was related (R° = 90-00, F = 26-89, p = 0-014) to the rainfall (X) in the growing season t (Equation 2). S, = -10-768 + 0-0343X, Equation 2 No relationship was found with temperature and flowering or seed pod production at this site. The green leaves of this plant tend to brown either just before or during flowering. The plant then enters a brief dormant phase. Plants are maintained during the dormant phase by means of a tuber. If the plant is pollinated seeds take at least six weeks to ripen (L. Farrell, pers. comm.) and mature from late July to late August or even the beginning of September (pers. obs.). No relationships between climate and the seed ripening phase and dormant phase were found. The seeds are approximately 130 x 30 x 30 um and are assumed to fly long distances. However a large number of seeds remain lodged in many pods so that when the flowering stalk falls over the seeds drop near to the parent plant. This has led to swarms of seedlings around parent plants in the years 1987-1995 at the two largest populations (Carey & Stewart in prep; L. Farrell, pers. comm. and G. Crompton, pers. comm.). DATASET A dataset was assembled which includes date of record, national grid reference, number of vegetative plants seen, number of flowering plants seen, habitat, and reason for disappearance (if appropriate). This dataset allowed the persistence of populations to be calculated. For many populations a record was not made in each year but it was often possible to make the assumption that the population persisted. In the case of two records at Box Hill in Surrey which are 106 years apart I, like Good (1936), have assumed that these are independent and demonstrate the recolonisation of a suitable site. There have been two populations at Burnham-on-Sea in Somerset but these are separated by approximately 8 km and 60 years and I have assumed that these records are not the same population. Unlike Good I have been more ready to accept that separate records at the same site are the same population. This is based on the demographic information gathered from seven populations, which indicates that there can be a long interval between generations and also that plants can remain in a less conspicuous state for many years. Both of these reasons would allow for the absence of records from a particular site. Records before 1895 are very widely spaced in time and make interpretation, especially of 356 P. D. CAREY figures, difficult. So for clarity in most of the Figures in this paper I have only analysed the period 1895-1998. Details of the earlier records are given by Good (1936). Although many of the records are accurate to 1 km some are only accurate to 10 km and this, along with the need to keep the location of some sites secret, means that maps of the distribution of sites in England presented in this study are based on a 10 km gnid. If it takes three years for a seedling to appear (Good 1936), and another three years to become mature (P. Carey & N. Stewart in prep.) a time-lag of about six years would exist between the climatic conditions which lead to establishment and the date of first record for a population (although there may have been some delay between a plant first flowering and the first year in which it was seen). In order to investigate the effect of climate on H. hircinum at the national level it was appropriate to use national weather statistics. Mean monthly rainfall data for England and Wales, supplied by P. Jones at the Climatic Research Unit at the University of East Anglia for the period 1895-1994, were available to compare with the records of H. hircinum. Equations | and 2 were applied to the rainfall data to give a probability of flowering and pod production for each growing season of the sequence. The product of these two equations gave a rough estimate (the standard error of this product based on the variance of Equations | and 2 was 0-54) of seed pod production per flowering plant in each year and, when multiplied by the number of populations recorded (assuming each population only had one flowering plant), an estimate of national seed pod production was obtained. The effect of severe droughts (more than 100 mm below the mean rainfall) on seeds was mimicked by reducing the seed pod production of year t, year t-1 and year t-2 in the model to zero. A six year time-lag was added to the date populations were recorded. Decadal smoothing was applied using LOWESS techniques (MINITAB 11) to both the data of the number of records and the estimate of seed production to mask some of the bias that is inevitable in the process of recording on a national scale. . RESULTS Reassessment of the records has led to a few discrepancies between the data presented here and those in Good (1936). There are additional records of which Good was not aware and a number of records that Good considered as geographically separate which I consider were not. The population recorded in north-western England at Ingleborough Hill in 1810.(herb. G. B. Woodruff) seems the most unlikely of any that was not documented by Good and I ask the readers of this paper to look on it with scepticism. The number of Himantoglossum hircinum populations in England varies dramatically with time (Fig. 1). There is a marked peak towards the end of the 1920s and early 1930s which is followed by a rapid decline and a remarkably constant number of populations from 1950—1990. Since 1990 there has been a rise in the number of populations. There is a high turnover of populations with a high proportion of new records and a high proportion of populations disappearing from the records (Fig. 1) throughout the period 1895-1998. Predictably, the mean age of populations declined as the number of populations increased (Fig. 2). Since 1940 the mean age of populations has increased steadily. For 439 of the 750 records of H. hircinum it is possible to infer the number of flowering plants present in the population. This provides data from at least one year for 116 of the 201 populations. The total size of the population is known for 193 out of the 750 records which represents at least one year from 39 different populations. The mean size of H. hircinum populations has not increased over the last 100 years but since 1945 two populations have become much larger than any other recorded in England. One population on a golf links in Kent produced over 3000 flowers in 1991, although few of these produced any seed pods. This population was an order of magnitude larger than the other substantial population in Cambridgeshire, which was itself an order of magnitude larger than any other population. The distribution of populations within England was mostly restricted to the south-eastern corner of England before 1910 with most of the populations being found around Dartford (Fig. 3a). During the next 30 years populations appeared in many of the counties of southern England (Fig. 3b). Despite the dramatic fall in the number of populations after 1940 the distribution of populations has not contracted (Fig. 3c) to the distribution that existed before 1910 (Fig. 3a). DISTRIBUTION AND ABUNDANCE OF HIMANTOGLOSSUM HIRCINUM She) ” S Oo — & mm} Q a Ge Oo aa an) ee © °@ e © 10 — x Ce ee ee ee ee To’ oo) fo) ° © © «88 © 68 © 680 esce = fo) fe) ry ® =) eso x & (e) z | + @gseesese oo) x 9° (oe) @) O x Cc o 8 Oo xO ® OOCx x x ex ®@ %® BC xo 8 soe OG g c 0 © xO C x@e xO XC @ BBB xxOKBO Ox CX KOs ooc © O : | ) : rr 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000 Year FicurE 1: The number of populations of Himantoglossum hircinum present in England each year since 1895 (solid circles), the number of populations that are first records (open circles) and the number of populations with a last record in that year (crosses). For clarity, only ycars where there was at least one record of a new population are shown and also only years where there is at least one last record are shown. 40 @ ® e @ e ® s { ) Oe e w Ae oa - e c @ S 30 — ae e. ® & A e ee? . = : ., {e) ® © Oo. ewe e® o * | re) ® S @ @ | le e CPi be | ® e e | ie?) 20) ee z oO a 3 4 ll 7) 8 8.5 9.5 10 10.5 eS) 12 12.5 13 Calyx length Ga 06 —Y Pi BU. gallii 0.35 + OU. minor 0.3 + > 0.25 + & $s Es 0.2 + & 0.15 + 0.1 T 0.05 | it 0 ae ze ci , 9.5 10.5 11 13.5 Keel length (mm) BU. Hien OU. minor ce > 2 oO =) > 75 8 8.5 9 9.5 10 10.5 11 LS: 12 Ne}S} 14 14.5 Wing length (mm) 035 0.3 + BU. gallii QU. minor 0.25 + — Opt 2 = 015 + 0.15 3 0.1 + 0.05 0 1.04 1.06 1.08 1.1 112 1.14 1.16 Ratio keel/wing length 1.02 370 F. KIRCHNER AND J. M. BULLOCK TABLE 1. THE DIFFERENCES BETWEEN ULEX GALLII AND U. MINOR (AS IDENTIFIED BY CHROMOSOME COUNTS) IN EACH CHARACTER WITH THE RESULTS OF T-TESTS. Ulex gallii Ulex minor Mean + se Mean + se i Primary spine length (mm) 19-0 + 0-36 10-7 + 0-25 19-04 Standard length (mm) 13-8 + 0-10 10-2 + 0-10 24-49 Calyx length (mm) 11-2 + 0-09 8-6 + 0-07 22:23 Keel length (mm) 11-9 + 0-08 9-1 + 0-07 26-35 Wings length (mm) 12-2 + 0-09 8-9 + 0-08 25-09 Keel/wing ratio 0-98 + 0-003 1-03 + 0-007 7-48 Means are calculated using the per bush means as samples. The variances of the two species differed in each comparison, so t-tests were carried out using separate variance estimates. In this method, the degrees of freedom are modified using the sample standard deviations of the two species. All tests were significant at P<0-0001 (1965) and Gloaguen (1986) reported that relative lengths of the wings and the keel can show some difference between U. gallii and U. minor. The ratio keel/wing length was included in the analyses, by calculating the ratio for each flower on a bush and taking the mean value per bush. Frequency histograms for primary spine, standard, calyx, keel and wing lengths of all plants sampled showed bimodal distributions (Fig. 1): These were formed by unimodal distributions of character values for each species (Fig. 1) and the peaks of each distribution appeared well separated between the species for each character. However, all distributions overlapped to a greater or lesser extent: 0-5 mm of overlap for standard and keel lengths, 1 mm of overlap for calyx and wing lengths, and 6-5 mm of overlap for primary spine lengths. The keel/wing ratio showed less of a species difference. Although the means were <1 for U. gallii and >1 for U. minor, as reported by Proctor (1965) and Gloaguen (1986), this character showed no clear bimodal distribution and substantial overlap (Fig. 1f). Despite these overlaps, t-tests showed significant differences between the two species in all six characters (Table 1). The overlaps in individual character values between the species means that identification based on single characters will always have some degree of error. The question is therefore, can a suite of characters be used to separate the two species completely and consistently? Discriminant analysis (Seber 1984) on all six characters was carried out using PROC DISCRIM in SAS (1990). This gave a good, but not perfect, discriminant function: all U. minor plants were classified correctly, but three U. gallii plants were misclassified (a success rate of 96-3%). This is illustrated in Fig. 2, which gives the graphic representation of the separation of the species based on the canonical discriminant functions. Most U. gallii plants have a score <0, but three have a score characteristic of U. minor plants, >0. Another way to address this question is to determine to what extent the measured ranges of each character can be used individually or in combination with other characters to separate the species. To do this we calculated which U. minor (or U. gallii) individuals had character values which fell within the ranges shown by the U. gallii (or U. minor) samples, for certain combinations of characters. The total of those individuals which would be classified as the wrong species using this method was used as a measure of the discriminating power of that character or group of characters. When carried out using the ranges measured by us at Gore Heath, no single character gave good discrimination between the species, but use of all characters together or all characters without spine length or keel/wing ratio accurately identified all but one of each species (Table 2). Proctor (1965) sampled plants from a large geographical spread of sites in Britain and his character values should be more representative of variation over Britain, so we repeated the analyses using Proctor’s (1965) character ranges. This gave worse results. Individual characters, especially spine and standard lengths, gave high proportions of misclassification, and combined characters still misclassified 15% of U. minor plants and 2% of U. gallii plants (Table 2). SEPARATION OF ULEX MINOR AND ULEX GALLII 371 25 Frequency Or 4 S BU. gallii OU. minor | “ol cal ci, Sat score ee: FiGurE 2. The distributions of the canonical discriminant scores for the 82 U. gallii plants and 46 U. minor plants, derived from the six measured characters. All U. minor scores are 0 apart from three misclassified plants. TABLE 2. THE RANGES OF CHARACTER VALUES FOR ULEX MINOR AND U. GALLII FROM THIS STUDY AND AS REPORTED BY PROCTOR (1965). Characters used Measures from this study Primary spine length (mm) Standard length (mm) Calyx length (mm) Keel length (mm) Wing length (mm) Keel/wing ratio All All except spine and ratio Measures from Proctor (1965) Primary spine length (mm) Standard length (mm) Calyx length (mm) Keel length (mm) Wing length (mm) All All except spine Ulex gallii Range Misclassified (%) 11-1-29-5 24 (29) 11-5—16-4 4 (5) 9-1-13-5 4 (5) 9-25-13-3 1 (1) 9-25-14-3 1 (1) 0-91-1-05 63 (77) ec) 1 (1) 8-34 79 (96) 10-5-18 5 (6) 8-5-14-5 15 (18) 9-15-5 3 (4) 9-5—15-5 5 (6) DY) 2 (2) Ulex minor Range Misclassified (%) 6:5-17-1 22 (48) 8-7-12-13 2 (4) 7:5—9-8 5) (WL) 7-7-10-0 19 (41) /7-5—-10-3 11 (24) 0-96-1-20 33172) 1 (2) 1 (2) 6-25 45 (98) 6-12-5 16 (35) 5-5-10-5 30 (65) 5-5-10-5 13 (28) 5-11 9 (20) 7S) 7 (15) The ranges of one or more characters for a species were used to classify plants of the second species as overlapping with the first species (i.e. misclassified), or not overlapping. The numbers of misclassified plants in each species are given. 372 F. KIRCHNER AND J. M. BULLOCK TABLE 3. INSECT SPECIES OBSERVED VISITING, AND PROBABLY POLLINATING, ULEX MINOR AND U. GALLI] FLOWERS. TOTAL NUMBERS OF OBSERVATIONS ALONG WITH THEIR RELATIVE FREQUENCIES ARE LISTED FOR EACH SPECIES. Ulex gallii Ulex minor Insect species number frequency number frequency Bumblebees Bombus terrestris/lucorum 28 0-364 16 0-356 Bombus humilis 5 0-065 4 0-089 Bees Andrena ovatula 5 0-065 M2 0-044 Apis mellifera 2 0-026 0 0 Hover-flies Sphaerophoria scripta 9 0-117 3 0-067 Syritta pipiens P)\ 0-273 19 0-422 Fristalis sp. 3 0-039 ] 0-022 Episyrphus sp. 4 0-052 0 0 PHENOLOGY The two species showed some differences in flowering phenology (Fig. 3). Phenology was examined in terms of the changes in mean number of flowers per plant (i.e. on the branches sampled) (Fig. 3a) and the proportion of plants in each census which reached their peak flower number at that census (Fig. 3b). The first U. gallii flowers were seen on 13 July, but U. minor started flowering later on 23 July. While the U. gallii population reached maxima in both the mean flower number and the proportion of plants at peak flower production on 18 August, these maxima were attained by the U. minor population on 9 September. Chi-square tests showed that the relative distribution of flower numbers between the censuses differed significantly between the species (y° = 557, df = 10, P<0-001), and the average date of the peak in flower number per bush was later for U. minor (median = 28 August) than U. gallii (median = 18 August) (Mann Whitney W = 4676, P<0-001). However, the species showed a large overlap in phenology as measured by the Proportional Similarity Index, Ps = 0-78. POLLINATOR OBSERVATIONS Five bee species were recorded visiting flowers (Table 3). Workers of the Buff-tailed and White-tailed Bumblebees Bombus terrestris and B. lucorum are difficult to separate with confidence, so we did not distinguish them in the field. Five hoverfly (Syrphidae, Diptera) species visited flowers (Table 2), but one, Paragus sp., was observed only once. Over the eight hours of observation, 77 insect visits were recorded for U. gallii, and 45 for U. minor (Table 3). Bombus terrestrislucorum and Syritta pipiens were by far the most frequent insects seen visiting Ulex flowers, with Bombus humilis, Andrena ovatula and Sphaerophoria scripta also frequent. Fisher’s Exact Test (used rather than a Chi-square test because of the small numbers of observations for several species) showed the pollinator communities of U. gallii and U. minor did not differ significantly (P = 0-546). This was illustrated by the high degree of overlap in the assemblages, as quantified by the Proportional Similarity Index, Ps = 0-83. DISCUSSION IDENTIFICATION OF ULEX SPECIES In accordance with other workers in Europe, we have assumed plants with chromosome counts of n = 32 to be U. gallii and with n = 16 to be U. minor (Alvarez Martinez et al. 1988; Bullock et al. 1998; Castroviejo & Valdes-Bermejo 1983; Fernandez Prieto et al. 1993). Using chromosome counts as an absolute method for distinguishing the two species, we were able to assess accurately the degree to which the species show overlaps in morphology and other traits. Other studies have been hampered by the circularity which results from assessing morphological overlaps using plants which have been identified to species using morphology. We found that the species showed clear differences in the average values of all the measured characters, with all apart from the keel/wing SEPARATION OF ULEX MINOR AND ULEX GALLII 373 a) SSSA OE Oh -- 1 --U. minor We 10 Mean flowers/plant = 25 oh 0 20 40 60 80 100 120 Days from first census (13 July) Ea U. gallii U. minor | 15 10 % of plants at peak flowering 0 105, 21 DO Ae Or AOR Doe. 10 80g iO 08) 120 Days from first census (13 July) Ficure 3. The flowering phenology of Ulex minor and U. gallii. a) Changes in the mean number of flowers per bush, with one standard error. b) The proportion of plants in the populations showing peaks in flower number at each census date. ratio showing bimodal distributions (Note: Proctor 1965 did not find a bimodal distribution for spine length). However, all characters showed some degree of overlap, and use of one character alone would give a minimum of 5% (using standard length) and a maximum of 77% (using ~ keel/wing ratio) of plants being misidentified. Using a suite of characters was more successful: discriminant analysis using all six characters gave only 2:-3% misclassification. More useful to field botanists is defining what suite of characters should be measured to best identify species accurately. Using the ranges in character values measured in this study, we found that the use of standard, calyx, keel and wing lengths resulted in misidentification of only two of the 128 plants, 1-6%. The 374 F. KIRCHNER AND J. M. BULLOCK keel/wing ratio is less useful for separating the two species. Proctor (1965) cast doubt on the idea that this ratio is always >1 for U. minor and <1 for U. gallii, although Gloaguen (1986) reiterated that this is a distinguishing feature. If this criterion is applied to the 128 plants sampled, 14 U. minor and 20 U. gallii plants are misclassified. Similarly, due to high plasticity, the length of the primary spine cannot be considered as a useful distinguishing character. We achieved 98-4% accuracy in identification using the character ranges of the plants measured on Gore Heath in Dorset. The ranges given by Proctor (1965) were from plants sampled from Plymouth in the west to Woking in east-central England, and so these better represent the variation in characters in England (although identification was not supported by chromosome counts). These gave a much worse result: when all four floral characters were used 7% of plants were misclassified. It is clear, therefore, that morphological characters cannot be relied upon to give a completely accurate separation of U. minor and U. gallii. Although we had a very high success rate using the character ranges measured at our study site, it is more relevant to use Proctor’s (1965) ranges when considering identification in Britain and Europe. Here, the 7% failure rate may sound small, but it translates to a large number of misidentifications if these character values are used extensively. Our plants were sampled from a mixed heath in an area of range overlap. The ranges of these species are largely disjunct (Bullock et al. 1998; Gloaguen 1986; Proctor 1965). In Britain U. minor is virtually confined to the south-east of England, whereas U. gallii occurs mostly in the west and north-west of England, Wales and the extreme south-west of Scotland. The distributions can be divided coarsely by a line running from Dorset to the Humber estuary, roughly halfway up the east coast. Therefore, it might be tempting to use geographical location as a distinguishing character (e.g. Cuba & Pardo 1997). However, this is a dangerous strategy and may lead to misidentification of plants occurring outside their recognised range limits. In this way responses to climate change, or other factors which may change plant distributions, may go undetected. There are several cases where the Ulex species are found well outside any simply-described range limits. U. gallii is found in a few locations in south-eastern England, most notably in Kent in the extreme south-east and, in large numbers, on the East Anglian coast. U. minor has some records in north Wales and on the south-western Scottish border (current distribution maps are held by the Natural Environment Research Council Biological Records Centre). It would be useful to check these records with chromosome counts from these disjunct populations. U. GALLII x U. MINOR HYBRIDS . The chromosome counts performed on the 135 plants did not provide any evidence for the occurrence of U. gallii x U. minor hybrids. These results — together with those of Bullock et al. (1998) who found a single putative hybrid (n = 24) out of 85 bushes sampled at Gore Heath — suggest that hybrids between U. gallii and U. minor are extremely uncommon. Therefore, the occurrence of hybridsation must be constrained in some way. INTERSPECIFIC POLLEN TRANSFERS. This study shows that the constraints on hybridisation are not flowering phenology or pollinator behaviour. It is an interesting and unexpected finding that U. minor had a significantly later start and peak in flower production than U. gallii (Fig. 3). Gloaguen (1986), working in Brittany, found that for U. minor flowering began in August, peaked in October (in terms of flower numbers) and finished in November. Flowering phenology differed between two U. gallii populations: flowering began in both during August, but one population peaked at the end of September and finished at the end of November, while the second peaked at the end of October and finished at the end of December. Therefore the species’ differences we found may not be repeated in other sites. Despite differences in start dates and peaks, the flowering periods of the two species overlapped to a large degree (Ps = 0-78) and there was high intraspecific vanation in the flowering period for both species (Fig. 3). Individuals of the two species bearing fully opened flowers simultaneously could be seen at any time during the flowering season. The insect species seen visiting the flowers of the two species of Ulex were the same, and the relative abundance of the insect species showed no significant difference between U. gallii and U. minor. It seems that exactly the same pollinator assemblage was visiting both Ulex species and that the insects were not distinguishing between the species. Given their overlapping phenologies and pollinator assemblages and the large degree of physical intermingling between plants, pollen SEPARATION OF ULEX MINOR AND ULEX GALLII 375 transfer between U. gallii and U. minor probably occurs frequently. Indeed, over the eight hours of observation, there was one definite sighting of a Bombus terrestrisNucorum worker moving from U. gallii to U. minor. Therefore, other factors may act as barriers to hybridisation between the two species. Alternatively, if hybrid seed is formed, the lack of hybrid plants may be caused by very poor germination or establishment of hybrids. Neither hypothesis has been explored fully. CONCLUSIONS Given the taxonomic difficulties within the section Neowilkommia of the genus Ulex, there is a need to use chromosome counts to investigate further the morphological and ecological correlates of groupings such as U. gallii, U. minor, U. europaeus, and U. europaeus x U. gallii. It is insufficient to rely on morphology alone to distinguish species or hybrids definitively. Other authors have reported the use of traits other than the gross morphological characters used in this paper, such as pollen grain size (Misset et al. 1982), epidermal structure (Godeau 1977), stoma size (Cuba & Pardo 1997) or isoenzymes (Misset & Fontenelle 1992). However, chromosome counts provide a discontinuous measure allowing definitive separation of species and hybrids. Problems may arise in cases where U. gallii plants appear to show 2n = 96 (e.g. Misset 1990; Misset & Gourret 1996), the same number as shown by U. europaeus. However, there is continuing controversy about such counts and there is a need to investigate ploidy levels in these species further. This study was carried out on one heath in Dorset. To expand this work and test the conclusions over the full geographic distribution of both species, the next stage should be to repeat the study at a range of sites over Britain, France and the Iberian peninsula. ACKNOWLEDGMENTS We thank George Else for identifying the insects, Ralph Clarke for statistical advice, Darren Edwards for help with the fieldwork and Nigel Webb and R. J. Gornall for comments on earlier drafts. REFERENCES ALVAREZ MARTINEZ, M. J., ef al. (1988). El genero Ulex en la Cornisa Cantabrica. I. Ulex gr. gallii-minor. Candollea 43: 483-497. BuLtock, J. M., Connor, J., CARRINGTON, S. & Epwarps, R. J. (1998). Chromosome numbers and flower sizes of Ulex minor Roth. and Ulex gallii Planch. in Dorset. Watsonia 22: 143-152. Carney, S. E., Hopces, S. A. & Arnoip, M. L. (1996). Effects of differential pollen-tube growth on hybridisation in the Louisiana irises. Evolution 50: 1871-1878. CastrRoviEJO, S. F. L. S. & VALDES-BERMEJO, E. (1990). On the identity of Ulex gallii Planchon (Leguminosae). Botanical journal of the Linnean Society 104: 303-308. CoLweLL, R. K. & Futuyma, D. J. (1971). On the measurement of niche breadth and overlap. Ecology 52: 567-576. CoriLuion, R. (1950). Phanérogames interéssantes pour la Bretagne. Bulletin de la Société scientifique de Bretagne 25: 131-140. Cusas, P. & Parpo, C. (1997). Correlations between chromosomal and morphological characters in subspecies of Ulex europaeus L. (Genisteae, Leguminosae) from the north-west of the Iberian Peninsula. Botanical Journal of the Linnean Society 125: 229-243. FERNANDEZ PRIETO, J. A., et al. (1993). Chromosome numbers and geographic distribution of Ulex gallii and U. minor (Leguminosae). Botanical journal of the Linnean Society 112: 43-49. GLOAGUEN, J. C. (1986). Les ajoncs de Bretagne. Bulletin de la Société botanique de France — Lettres botanique 133: 363-385. Gopeau, M. (1977). Observation au microscope électronique a balayage de |’épiderme séminal d’Ulex europeaus L., U. minor Roth, U. gallii Planchon. Bulletin Société science naturelle Ouest de la France 75: 83-89. LaMBINON, J. (1962). Note sur les Ulex du Massif Armoricain. Lejeunia 9: 64-66. LEEBENS-Mack, J. & MILLIGAN, B. G. (1998). Pollination biology in hybridising Baptista (Fabaceae) populations. American journal of botany 85: 500-507. 376 F. KIRCHNER AND J. M. BULLOCK Misset, M. T. (1990). Données caryologiques chez le genre Ulex L. (Papilionidae) dans le Massif Armoricain. Taxon 39: 630-635. Misset, M. T. & FONTENELLE, C. (1992). Protein relationships between natural populations of Ulex europaeus and U. gallii (Faboideae, Genisteae) and their hybrids. Plant systematics and evolution 179: 19-25. Misset, M. T. & Gourret, J.-P. (1996). Flow cytometric analysis of the different ploidy levels observed in the genus Ulex L Faboideae-Genisteae in Brittany (France). Acta Botanica 109: 72-79. MisseT, M. T., GourreT, J.-P. & Huon, A. (1982). Le pollen d’Ulex L. (Papilionoideae): morphologie des grains et structure de l’exine. Pollen et spores 24: 369-395. Proctor, M. C. F. (1965). The distinguishing characteristics and geographical distributions of Ulex minor and Ulex gallit. Watsonia 6: 177-187. Proctor, M., YEO, P. & Lack, A. (1996). The natural history of pollination. Bath Press, Bath. Rozzi, R., ARRoYo, M. K. & Armesto, J. J. (1997). Ecological factors affecting gene flow between populations of Anarthrophyllum cumingti (Papilionaceae) growing on equatorial- and polar-facing slopes in the Andes of Central Chile. Plant ecology 132: 171-179. SAS (1990). SAS/STAT users guide. SAS Institute, Carey, North Carolina. SEBER, G. A. F. (1984). Multivariate observations. John Wiley & Sons, New York. WEIBLEN, G. D. & BREHM, B. G. (1996). Reproductive strategies and barriers to hybridisation between Tellima grandiflora and Tolmeia menziesii (Saxifragaceae). American journal of botany 83: 910-918. (Accepted December 1998) Watsonia 22: 377-385 (1999) 87/1 Use of herbarium material for mapping the distribution of Erophila (Brassicaceae) taxa sensu Filfilan & Elkington in Britain and Ireland ti=CyGeRICH Department of Biodiversity and Systematic Biology, National Museum and Gallery, Cardiff CF1 3NP and J. LEWIS 19 Wingfield Close, The Common, Pontypridd CF 37 4AB ABSTRACT Specimens held in herbaria are a major botanical resource in Britain and Ireland but are currently being under-utilised by botanists in Britain and Ireland. The distribution of the taxa within the genus Erophila in Britain and Ireland has been selected as an example to illustrate the importance of herbarium specimens and one use for the collections. Keyworbs: Erophila verna, Erophila majuscula, Erophila glabrescens. INTRODUCTION Herbarium specimens are a major botanical resource which can be used for a range of purposes including taxonomic research, verification of identification, to provide geographical, historical and ecological information, and as an educational resource. The long history of collecting in Britain and Ireland means that a wealth of plant material is held in publicly accessible collections in Museums, Universities and other institutions around the British Isles. Herbaria are currently being under-utilised by botanists in Britain and Ireland. For example, Rich & Sydes (1999) investigated two species included in the Scarce Plants Project and were able to trace 209 records in nine major herbaria additional to the 40 records in the Scarce Plants Database. The under-utilisation of herbaria may be a function of the increasing quality of books and illustrations allowing identification without comparison against a set of reference material, and the change in attitude to collecting related to concern for conservation of the plants. It may also simply be due to botanists, being unaware of the importance of existing collections or that they are accessible to the public. Limited public spending and increasing financial pressures have meant that it is becoming increasingly difficult to justify spending money on maintaining and enhancing TABLE 1. NUMBER (AND %) OF RECORDS OF EROPHILA SPECIES FROM DIFFERENT PARTS OF THE BRITISH ISLES Area E. majuscula E. verna E. glabrescens Southern England (v.cc. 1-32) 65 (50%) 656 (48:5%) 59 (22%) Northern England (v.cc. 33, 34, 36-40, 53-70) 24 (18:5%) 320 (23-5%) 84 (31-5%) Wales 11 (8-5%) 116 (8-5%) 21 (8%) Scotland 17 (13%) 122 (9%) 61 (23%) Ireland 13 (10%) 128 (9-5%) 39 (14.5%) Channel Islands 0 (0%) 11 (0-8%) 2 (0:7%) 378 T. C. G. RICH AND J. LEWIS Ficure |. Distribution of all Erophila records held in the Fielding Druce herbarium, University of Oxford (OXF), showing the strong local representation. herbaria which are not being used. Some financial controllers have even questioned the value of keeping herbaria at all. The distribution of the taxa within the genus Erophila in Britain and Ireland has been selected as an example to show one role that herbarium collections can play in botanical research, as one of a range of issues being considered by the Herbarium Managers’ Group of the UK Systematics Forum. The UK Systematics Forum was set up in 1994 to promote co-ordination and communication between the major UK collections-holding institutions and the wider systematics community, and a national strategy for systematic biology research has been drawn up in which zoological and botanical collections play a major role. Filfilan (1984; summarised in Filfilan & Elkington 1988, 1998, and Elkington 1991) carried out a cytotaxonomic study of Erophila populations in Britain, the results of which were correlated with those of Winge (1940). Three groups of cytotypes were distinguished at the specific level: Erophila majuscula Jordan, a densely hairy diploid, FE. verna (L.) Chevallier sensu stricto, a medium polyploid, and E. glabrescens Jordan, a sparsely hairy to glabrous high polyploid. The distributions were summarised by vice-county in Filfilan & Elkington (1998), but no distribution maps were presented. Hectad (10 km x 10 km square) maps of these species are required for the B.S.B.I. Atlas 2000 project, but relatively few records have been verified to species level. The opportunity was taken to draw together records for these species in selected major herbaria to show the wealth of information available. MAPPING EROPHILA TAXA 379 Figure 2. Distribution of Erophila majuscula based primarily on herbarium records. METHODS Data on specimens were abstracted onto standard “pink cards” from material held in the following herbaria (abbreviations follow Kent & Allen 1984): BEL (74 sheets), BM (400 sheets, c. half the material), DBN (145 sheets, mostly named by Elkington & Filfilan), E (c. 200 sheets, partly named by Elkington & Filfilan), K (129 sheets, c. half the material), LANC (25 sheets, Cumbria material determined by T. T. Elkington only), LIV (155 sheets), MANCH (157 sheets), NM'W (161 sheets, partly named by Elkington & Filfilan), OXF (189 sheets, many with up to 5 or 6 collections on each sheet), RNG (114 sheets) and TCD (32 sheets). Identifications by T. T. Elkington and/or S. A. Filfilan were accepted directly, with other material named by T. C. G. Rich. Grid references were allocated to the most appropriate hectad (10 km x 10 km square) following standard B.R.C. practice, though in some cases the original locations were not known with certainty (e.g. “near Manchester’; LIV). The handwriting on some sheets was difficult to read, and minor discrepancies © were encountered between apparent duplicates. The labelling on some sheets is so poor (e.g. Druce’s material in OXF) that some records were not included to avoid errors. Other records with illegible handwriting, or where the locality was ambiguous or could not be traced, have not been included on the maps. 380 T. C. G. RICH AND J. LEWIS ig ¥ er, trai f | ok | eee re Ficure 3. Distribution of Erophila verna sensu stricto based primarily on herbarium records. A few additional records published in the literature and other records held by the Biological Records Centre (BRC), Monks Wood have been included, but no systematic search for other records has been carried out so that the maps very largely represent the data held in herbaria. The pink cards have been deposited at BRC, to which requests for details of the records should be addressed. RESULTS The exercise in collating records also provided an opportunity to test the taxonomy, which on the whole was found to work well. E. majuscula is a distinct species with a dense rosette (due to the short petiole), dense hairs and obovate fruits. In most cases identifications were checked by measuring seed sizes. E. glabrescens by comparison is sometimes difficult to separate from E. verna, and the dividing line between the two on sparsely hairy plants would appear to be somewhat arbitrary. The extent to which the petals are bifid does not appear to correlate well with hairiness and has not been used by TCGR. On mixed sheets, plants allocated to this species were almost always the smallest, and one wonders if hairiness is somewhat dependent on growth/habitat. A strict view of E. glabrescens has been taken to include only the most sparsely hairy plants. Most material was referable to the variable FE. verna sensu stricto, of which a broad view has been taken. MAPPING EROPHILA TAXA 381 Figure 4. Distribution of Erophila glabrescens based primarily on herbarium records. The correspondence between different herbaria and with specimens named by Filfilan and Elkington was found to be good, though it would not be too surprising if some duplicate collections in different herbaria have been given different names. It was not possible to name some immature or senescent material. Mixed collections do occur quite regularly, matching the situation in the field. Some Erophila names have been used historically in a different sense to how they are used by Filfilan & Elkington, and it is not possible to use old names, with the possible exception of material named as E. virescens Jordan by E. S. Marshall, which is usually E. majuscula. The herbaria were found to have very strong representations of local material, with patchy representation of specimens from elsewhere. For example, Fig. 1 shows the distribution of all Erophila records held in OXF, which, as might be expected, shows a strong concentration from Oxfordshire. The maps for the three species are shown in Figs 2-4. Records for E. verna s.s. var. praecox (Steven) Diklic were also abstracted from some, but not all, herbaria, and are shown in Fig. 5. The numbers of records from various parts of the British Isles are summarised in Table 1. The numbers of distinct collections per decade (i.e. excluding duplicates) are summarised for all three species in Fig. 6. The changes with time reflect variation in collecting effort rather than changes in frequency of the plants (Rich 1997). There is a peak in collecting around the turn of the century when botanists such as E. S. Marshall and J. E. Little collected large quantities of material. Most post-1980 records are from Cumbria, Monmouthshire and Ireland. 382 T. C. G. RICH AND J. LEWIS Ficure 5. Distribution of Erophila verna var. praecox based on herbarium records. DISCUSSION The maps show how the large number of specimens held in UK and Irish collections can be used to provide valuable information on distribution. With any taxonomic revision there is an immediate requirement for information to bring the data up to a comparable standard with other taxa, and the collections provide an efficient means of doing this. Other taxa in the British Isles for which additional data are required for the Atlas 2000 project include the Juncus bufonius, Utricularia intermedia and Luzula multiflora aggregates. Whilst perhaps up to 20% of the Erophila material in different herbaria is duplicated (often material distributed through Botanical Exchange clubs), the majority is unique and often shows a strong local representation. For instance, there is much material from Edinburgh in E, from the Merseyside area in LIV and from Glamorgan in NMW. It is thus important to examine material from a range of herbaria to obtain a general overview of the distribution. However, a comparison of the summary map of all records (Fig. 7) with that from the Atlas of the British flora (Fig. 8) shows that some areas are still under-represented in herbaria compared to the frequency in the field, and caution is required in drawing firm conclusions about distribution or frequency based on herbarium material alone. Nonetheless, the information is invaluable for indicating which species may be present in an area and hence should be searched for. MAPPING EROPHILA TAXA 200 ay ao oO Number of records ro) oO 50 1800 1850 1900 1950 2000 Decade Ficure 6. Number of records per decade for Erophila majuscula (OQ), E. verna (@) and E. glabrescens (xX) Figure 7. Distribution of Erophila verna sensu lato combined from Figs 1-5. 384 T. C. G. RICH AND J. LEWIS VERNA agg. , mf h Ficure 8. Distribution of Erophila verna sensu lato (Perring & Walters 1962). The maps show that FE. majuscula is the least common species, collected most frequently in southern Britain (Table 1) and becoming scattered or rare elsewhere. It is probably sufficiently uncommon to qualify as a Nationally Scarce species. E. glabrescens is not common but is widespread, and may be the commonest taxon in the north and west of both Britain and Ireland. E. verna S.S. 18 the commonest and most widespread species. Examination of the herbarium material has also resulted in additional records for many Vice- counties to be found. The lists of Vice-counties for which we have seen specimens of each species are as follows (78 records new to the list of Vice-counties given in Filfilan & Elkington 1998 are given in bold): E. majuscula: V.c. 1, 5,6, 7,9, 11, 12, 135 14, 16,17, 18, 20; 215 22,23, 2652723) sles 2ao ooo 38, 40, 41, 49, 50, 51, 55, 57, 58, 59, 62, 64, 65, 68, 78, 80, 82, 83, 85, 90, 95, H9, H15, H17, H21, H22, H37, H38 and H40. E. verna: V.c. 1, 2, 3, 4, 5;6; 7 8; 9, 10) th 12, 13; 14) 15, 16,17, 1819520) 2 oe ee 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59; 60; 62; 64, 65, 66; 67; 68, 69) 70; 12773, 16; 77, Veo Omse. 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 94, 95, 96, 97, 98, 99, 100, 103, 105, 106, 107, 108, 109, 110, 111, H1, H2, H3, H4, H5, H6, H7, H8, H9, H10, H11, H12, H13, H15, H16, H17, H18, FN H20 2 He 2) 23 hi24 25. H26, H30, H31, H33, H35, H36, H37, H38, H39 and H40. MAPPING EROPHILA TAXA 385 Pealdarescensy Cale s9.0.6,. 9) Oss I2IS 145, 16, 17, 19; 20;21, 23, 24, 26, 27, 28, 29, 30, 31, 33, 34, 35, 36, 37, 38, 40, 41, 42, 43, 44, 46, 47, 49, 51, 52, 54, 55, 56,57 58, 59, 60, 62, 63, 64, 65, 66, 69, 70, 72, 73, 74, 76, 77, 78, 79, 80, 81, 82, 83, 85, 88, 89, 90, 94, 95, 96, 98, 100, 101, 106, 107, 108, 109, 110, 111, H6, H8, H9, H11, H15, H16, H17, H20, H21, H23, H27, H28, H30, H34, H36, H38, H39 and S. Filfilan & Elkington (1998) suggested that the map of E. verna subsp. spathulata (Lang) Walters in Perring (1968) was applicable to their var. praecox. However, a number of the populations in the west with short fruits are E. glabrescens rather than E. verna; Perring’s (1968) map is therefore not completely consistent with the modern treatment. The different distribution of this variety suggests that there may be some merit in retaining it as a taxon, though what the ecological basis for the difference in distribution is requires investigation (e.g. van Andel et al. 1986). The importance of herbarium material to botanical research must be stressed to the B.S.B.I. membership. Whilst many herbaria only lend specimens to recognised institutions (e.g. local Museums), and some do not have the resources to post material, in most cases requests to visit or borrow material will be welcomed. ACKNOWLEDGMENTS We would like to thank the Keepers of the above herbaria for access to or loan of specimens, and John Edmondson, Trevor Elkington, Paul Hackney, Rob Huxley, Stephen Jury, Serena Marner, John Parnell, Berni Shine, Sue Zmartzti, Donald Synnott and Elizabeth Woodgyer for their help. Luke Corfield, Rhiannon Evans and Michael Lister helped to abstract records during their work placements. The maps were plotted using DMAPW software by Alan Morton. REFERENCES ELKINGTON, T. T. (1988). Erophila, in Ricu, T. C. G. (1991). Crucifers of Great Britain and Ireland. B.S.B.1. Handbook No. 6. B.S.B.I., London. FILFILAN, S. A. & ELKINGTON, T. T. (1988). Erophila, in Ric, T. C. G. & Ricu, M. D. B. (1988). Plant crib. B.S.B.I., London. FILFILAN, S. A. & ELKINGTON, T. T. (1998). Erophila, in Ricu, T. C. G. & Jermy, A. C. (1998). Plant crib 1998. B.S.B.I., London. FILFILAN, S. A. (1984). Cytology and biosystematics of Erophila DC. (Cruciferae). Ph.D. Thesis, University of Sheffield. Kent, D. H. & ALLEN, D. E. (1984). British and Irish herbaria. B.S.B.I., London. PERRING, F. H.. & WALTERS, S. M. (1962). Atlas of the British flora. Thomas Nelson, London. PERRING, F. H. (1968). Critical supplement to the Atlas of the British flora. Thomas Nelson, London. Ricu, T. C. G. (1997). Using botanical records to interpret changes in frequency of British plants. The biology curator 10: 8-12. Ricu, T. C. G. & Sybes, C. (in prep). Scarce Plants in Britain: an independent assessment of the database. Watsonia. VAN ANDEL, J., Rozun, N. A. M. G., Ernst, W. H. O. & NELIsseN, H. J. M. (1986). Variability in growth and reproduction in F, - families of an Erophila verna population. Oecologia 69: 79-85. Wince, O. (1940). Taxonomic and evolutionary studies in Erophila based on cytogenetic investigations. Compt. Rend. Trav. Lab. Carlsberg (Sér. Physiol.) 23: 41-74. (Accepted January 1999) reels oan ae Cao ol ens ae Oaat Watsonia 22: 387-395 (1999) 387 The history and distribution of Phyteuma spicatum L. (Campanulaceae) in Britain B. R. WHEELER* and M. J. HUTCHINGS School of Biological Sciences, University of Sussex, Falmer, Brighton, Sussex, BN] 9QG ABSTRACT Phyteuma spicatum L. (Campanulaceae) is a rare species in Britain occurring in low numbers in a small area of East Sussex (v.c. 14). Its history has been investigated, and herbarium records and Flora accounts examined, to determine whether it was formerly more widespread and abundant in Britain. It appears that P. spicatum has always been confined to East Sussex, but that it was more widespread within this vice-county in the past. Anecdotal information also suggests that it was more abundant at the end of the 19th century and during the early 20th century than it is at present. KEYWORDS: Spiked rampion, rare plant, current distribution, historical distribution, England. INTRODUCTION Thirty species from the family Campanulaceae occur in Britain as natives, introductions or garden escapes (Stace 1997). Three of these species belong to the genus Phyteuma. P. orbiculare L. (Round-headed Rampion) is native and is found in chalk grasslands from Wiltshire to East Sussex. P. scheuchzeri All. (Oxford Rampion) is an introduced species which was formerly on walls and pavements in Oxford and is now naturalised in limestone cracks at Inchnadamph, W. Sutherland (Stace 1997). The third member of the genus, P. spicatum (Spiked Rampion), is confined to East Sussex (v.c. 14). According to Damboldt (1976) there are two subspecies of P. spicatum. Subspecies spicatum has creamy-white flowers, whereas those of subsp. coeruleum are blue. Subspecies spicatum is the subspecies found in all the East Sussex sites, and is regarded by many as a native of Bnitain (Hall 1980; Clapham et al. 1987). Garden escapes are usually blue-flowered (Stace 1997). Whereas Bentham & Hooker (1954) place a question mark after “native” in their flora entry for P. spicatum, and Grigson (1958) fails to mention the species at all, Hall (1980), Clapham et al. (1987) and Stace (1997) all regard it as native. Its inclusion in the second and third editions of the British Red Data Book of Vascular Plants (Perring & Farrell 1983; Wigginton 1999), and its addition in 1992 to Schedule Eight, under Section 13 of the Wildlife and Countryside Act of 1981, support this view. P. spicatum is found very locally on roadside verges, on steep roadside banks and in woodlands, on acid soils of the Ashdown Beds and Wealden Clay. The communities in which it is found on roadsides correspond to either the W25a Rubus fruticosus agg.—Pteridium aquilinum underscrub, Hyacinthoides non-scripta sub-community or, in more open areas, the Arrhenatherum elatius grassland, Centaurea nigra sub-community (MGle), of the National Vegetation Classification. In wooded sites it is a component of Quercus robur-Pteridium aquilinum-Rubus fruticosus woodland (W10), particularly the Anemone nemorosa sub-community (W10b) (Rodwell 1991, 1992; Whceler 1997). P. spicatum has also been recorded as a garden escape in Warwickshire, Staffordshire, Merionethshire, Derbyshire and Roxburghshire (Pernng & Farrell 1983). Outside the British Isles, P. spicatum is endemic to Europe, occurring in many Central and Atlantic European countries. Ellenberg (1988) names P. spicatum as a component of the “noble broadleaved wood” (Fageztalia), lime and oak-hormnbeam woods, silver fir woods, woodland margins, and montane and sub-alpine * Address for correspondence and reprint requests: 2 Lodge Cottages, Wych Cross Place, Wych Cross, Forest Row, East Sussex, RH18 5JJ 388 B. R. WHEELER AND M. J. HUTCHINGS meadows. Full details of the communities in which the species occurs, and of its ecology, are to be presented in a Biological Flora account (Wheeler & Hutchings, in prep.). Willems (1980) has studied the size of continental populations of P. spicatum in the Leudal area of the Netherlands, a country where the species is rare. The number of plants he recorded declined between 1965 and 1980, culminating in population extinction. Detailed study of the demographic behaviour and changes in numbers of plants in three of the remaining British populations of P. spicatum will be presented in a future paper (Wheeler & Hutchings, in prep.). The present paper analyses the past distribution and abundance of P. spicatum in Britain to determine whether its current limited distribution and rarity has been typical throughout its history, or whether it has suffered a contraction in range and abundance. Evidence for such a contraction would suggest that development of a programme of active management is urgently needed to conserve the species. METHODS The current and historical distribution and abundance of P. spicatum in Britain were analysed using historical records, manuscripts, herbarium specimens and field observations. Records and specimens were examined from herbaria at BM, BTN, CGE, LIV, K and NMW. Records from other herbaria and site locations were supplied by English Nature, Sussex and Surrey Team. The B.S.B.I. vice-county Recorder was consulted for known locations, and the Joint Nature Conservation Committee and the Biological Records Centre at the Institute of Terrestrial Ecology, Abbots Ripton, supplied information on historical locations. In addition, known locations at which P. spicatum still occurs, and sites at which it has been recorded but at which its presence was unconfirmed between 1986 and 1996, were searched for the species. All records of P. spicatum are listed in the Appendix. THE HISTORY OF PHYTEUMA SPICATUM IN THE BRITISH ISLES The earliest account of P. spicatum in the British Isles is in Gerarde’s ‘Herball’ (1597) under Rapuntium maius, an early name for the species. It was described as a garden plant, grown for its culinary and medical properties (the root was boiled and eaten in salads), rather than as a wild species. Gerarde stated that the plant was also known as Rapunculum alopecuron because of “‘the eare or spike that is full of flowers, which are like almost to tailes of foxes when they are bloomed”’. Gerarde’s plant was blue-flowered, “‘sometimes white or sometimes purple’, indicating that P. spicatum grown in gardens in Britain at that time was subsp. coeruleum R. Schulz (Damboldt 1976), which now grows wild in mainland Europe but only grows as a garden escape in Britain. Gerarde (1633) enlarged the list of names given to the species to include Alopecuri comoso flore, Rapunculum sylvestre and Rapunculus sylvestris spicatus. Parkinson (1640) also described the plant under the name Rapunculus spicatus Alopecuroides, or the “Long Foxtaile Rampion”. He stated that as well as being cultivated as a culinary plant, the species grew wild in “divers places of this land’. This is the first record of P. spicatum as a wild species in Britain. Parkinson also commented on the origin of the name, stating that Rapunculus was a diminutive of Rapum (the turnip) and that the species was so-called because of the similarity in appearance and edibility between its root and the vegetable. Further names included Rapum sylvestre majus, Rapunculum alopecuron, Rapunculus Alopecuroides longa spica and Rapunculus nemorosus primus. The first specific location given for P. spicatum as a wild plant is Knight’s Farm, Mayfield, East Sussex in 1825. This record forms part of William Borrer’s herbarium, now housed in K. Borrer’s botanical knowledge of Sussex was unequalled in his day (Arnold 1907; Wolley-Dod 1937), lending credence to this record. A specimen from this first record, sent to the Linnean Society “was not considered an English plant” (Branwell 1872). From 1825 onwards, there are many records and herbarium specimens of the species. The most extensive herbarium collection is in BM. Most specimens in other herbaria were duplicates of these. All the documented locations of P. spicatum growing wild were in East Sussex, and there is no evidence of the species as a wild plant in other counties — only as a garden plant or nearby escape. Most herbarium records provide information on the abundance and distribution of the species. Although some records are unspecific, referring only to a particular parish, the names of PHYTEUMA SPICATUM IN BRITAIN 389 particular farms or woods are frequently given. A note by W. A. Bromfield, made in 1835 (herbarium of W. B. Hemsley, Booth Museum, Brighton), 1s typical of the accuracy and form of these records; “‘...1n the middle of a wood on Hole Farm in Buxted Parish, five or six miles from Uckfield; one of many stations about Mayfield and Waldron where this plant occurs, and which is certainly indigenous though so long overlooked by British botanists”. Bromfield’s view of P. spicatum as indigenous 1s supported by Jenner’s (1845) observation on the species in woods at Warbleton where “No one who saw it there could doubt its being native”. In 1907, Arnold suggested that specimens of P. spicatum found in the wild were escapes from Warbleton or Michelham Priories. This theory originates from Michelham Priory itself. It was still in the apothecaries’ garden at Michelham Priory in the late 1980s (FitzGerald 1987), and an exhibit in the Priory museum claimed that it was “brought to Michelham by the canons from its native France’, although no evidence is given to support this claim. The number and size of populations of P. spicatum in East Sussex appears to have been much larger in the past than at present. Throughout the 19th century, herbarium specimens were consistently accompanied by notes referring to its great abundance. In 1835, Bromfield “met this plant in plenty” (BTN) and Jenner’s (1945) account of the species states that it “grows in the woods in such abundance ...’”’. A later extract from a letter to W. B. Hemsley in 1875 (Herb. Borrer., K) describes “Abbots Wood which is quite carpeted with Phyteuma spicatum. We were there for two hours and which ever way we walked they were as common as dandelions”. In 1996, Abbots Wood supported fewer than 30 plants. Several herbarium specimens from the early part of the 20th century also refer to P. spicatum as an abundant species. For example, Bray in 1919 (FitzGerald 1988) stated that “it came from Bramble Grove Wood ... There is a good bit of it.” However later records suggest a decline. Druce (1932) described the species as rare, and other records around this time cease to describe the plant as occurring in large numbers. Wolley-Dod (1937) records it in Abbots Wood as “formerly abundant, now much less so”. Later records are more specific about the location of the plant, referring to populations in small areas of the wood, rather than throughout. For example, Richards (1942 [in Herb. Babington, CGE]) records it as in a “Felled portion of Abbots Wood near Milton Hide”’. DISTRIBUTION OF PHYTEUMA SPICATUM IN EAST SUSSEX FROM 1825 TO 1996 Figures la—e show the presence of P. spicatum within tetrads in East Sussex from 1825, the year of the first confirmed record of the species, to 1996. Between 1825 and 1875 (Fig. 1a) most records were from the Heathfield area. They do not support the suggestion that wild populations of P. spicatum originated from Michelham Pnory (see above), since only one tetrad near the Priory (TQ/5.0T) has records of the species, whereas there are many tetrad records 10 km further north. The location of Warbleton Priory (TQ/6.1P) is closer to the early records and may be more credible as a possible source of P. spicatum, if the species is not native. The distribution of P. spicatum appeared to spread between 1875 and 1925 (Fig. 1b), although few records remained in the tetrads in which the species had been previously recorded. This apparent increase in the geographical range of the species may be explained by the increase in regionally specific botanical documents during this period, allowing greater detail and accuracy in recording (Hemsley 1875; Arnold 1887, 1907), and by the increased number of amateur botanists collecting specimens for herbaria. The majority of herbarium specimens of P. spicatum originate from this period. The absence of records from many of the earlier locations cannot be explained. Their recurrence in later decades suggests a lack of recording in these locations, rather than that P. spicatum had become locally extinct. The records of P. spicatum reached their greatest abundance between 1926 and 1975 (Fig. Ic), but even during this period only 18 tetrads contained records. P. spicatum was still confined to a small area within a circle of radius 9 km, with two areas of concentration. These were the Heathfield/Cross-in-Hand area (TQ/5.2) and the woodland in and around Abbots Wood (TQ/5.0). From 1975 to 1989, the range of P. spicatum diminished (Fig. 1d), shrinking back to the two stronghold areas where it had formerly been concentrated. The outlying tetrad records shown in Figures 1b and c had been lost. By 1996 the strongholds for P. spicatum had declined in size (Fig. 1e), 390 B. R. WHEELER AND M. J. HUTCHINGS A emme ee aa CEC eae CORRES Baa Seaeeeeee ago ES! 2a eeee8 BE en ERE Oe ES (88 SESE CRESS BEE OR Pee eee eee aes eee eR ie" I eee SRGH) | BERER BREE. SS ORR Eee eee eee EEaEe BERS BER eee S08 0008 Gone on) os Soe Coe ee ES 68 SEER CESS Bees eee 1c ane, come Ee =? Ie eee Baan aan 0 ee ES (88 SERRE PEs BES Bees Pees eee JE EB 8 BEES Eee Ae (20000000 06008 S008 CS eee J| SSS888 28 = eR EE 162 =" IS eee Gane( Range Genes -oaee eee eee —H ee FiGurE la-e. The distribution of P. spicatum during the period 1825 to 1996, plotted from the historical and extant site records. (a) 1825-1875; (b) 1876-1925; (c) 1926-1975; (d) 1976-1989; (e) 1996-—present. Distribution is shown as presence within 2 x 2 km square tetrads in v.c. 14, East Sussex. Those tetrads in which presence is marked as an approximate location only are an attempt by the author to place the earliest (1825 to 1875) historical records of P. spicatum. These early records are open to interpretation of location, due to the vagueness of descriptions. PHYTEUMA SPICATUM IN BRITAIN 391 TABLE 1. THE NUMBER OF PHYTEUMA SPICATUM PLANTS AT EACH EXTANT SITE DURING THREE RECENT CENSUSES, SHOWING THE PERCENTAGE CONTRIBUTION WHICH PLANTS AT EACH SITE MAKE TO THE TOTAL POPULATION IN BRITAIN. Sites 1980/1 1986/7 1995/6 total % of total % of total % of number of British number of British number of British plants population plants population plants population Pound/Bridgelands Farm 2 <1 0 0 0 0 Brown’s Lane 5) II 13 a 14 ] <1 Little England Farm 50 3) Dil i >) l Malls Bank Farm 16 4 18 5) 7] 2 Tinkers Lane 198 50 175 44 285 83 Abbots Wood sites VY 19 30 8 23 7 Dower House _ — 7 v ? u Maynards Green ~ — 78 20 14 + New Sapperton Farm ~ - 12 30 ? y Rushlake Green - = - - 4 ] St Dunstan’s Farm - - - - 4 ] Total 397) 395 343 ? Denotes no available information on population numbers during the survey period; — denotes no plants known at the site during the survey period. leaving only nine tetrads containing records. The appearance of one new tetrad supporting a colony at TQ/5.0N (Abbots Wood) is misleading, since this record, which is adjacent to the older record of TQ/5.0T, is due to improvement in the accuracy of an earlier record. Both TQ/5.0N and T are a single record from one woodland which supports a handful of plants in a very small area crossing the tetrad boundary line (see Appendix records). An analysis of the recent and current distribution of P. spicatum in Britain is given in Table 1, which shows, for three census dates, plant numbers, and the percentage of the total number of plants of P. spicatum contributed by populations at each extant site. The 1980/1 census was undertaken by the county recorder of the B.S.B.I. for the Sussex Botanical Recording Society, the 1986/7 census was commissioned by English Nature, and the 1995/6 census was carried out during this study. In 1987 there were eight sites with extant populations, supporting almost 400 plants (FitzGerald 1988). This was a similar number of plants to that recorded in 1980/1, but the distribution across sites had changed between the two dates. The census in 1996 confirmed the persistence of P. spicatum in seven of the tetrads occupied in 1987. The eighth had become inaccessible, so that the continued presence of P. spicatum here remained unconfirmed. P. spicatum was also discovered at two new sites. However, with the exception of one site, the numbers of plants at the sites where the species survived were much reduced since the 1987 census (Table 1), and the total number of plants had declined by 13% between 1987 and 1996. Although the spread of plants across the Bnitish sites was uneven at all three census dates referred to above, it has recently become more heavily concentrated on just one site, emphasizing the importance not only of conserving the population at this site but also of increasing the size of P. spicatum populations at the other remaining sites in which it survives. A catastrophic event or inappropriate management at the site of its largest population could now destroy over 83% of Britain’s P. spicatum plants. CONCLUSIONS Historical records show that, although P. spicatum has never been widely distributed in the Bnitish Isles, it was formerly more widespread, and occurred in much larger numbers, than it does now. The low numbers of plants which now remain warrant its legal protection under Schedule Eight, 392 B. R. WHEELER AND M. J. HUTCHINGS Section 13 of the Wildlife and Countryside Act (1981), and there can be little doubt that an effective plan for management of the sites where it still survives is now vital to safeguard the species against the risk of extinction in the Bnush Isles. ACKNOWLEDGMENTS We thank the following for providing records and/or information on sites; English Nature (Sussex and Surrey Team), Forestry Commission (Weald Forest District), Peter Davys, Desmond Gunner, Paul Harmes (and the Botanical Society of the British Isles), A. W. Jones, Elizabeth Rich, Dr Tim Rich, Dr Francis Rose, Janet Simes, Matthew Thomas; and the Keepers of BM, BTN, CGE, K, LIV, NMW and RNG for access to specimens. REFERENCES ARNOLD, F. H. (1907). Flora of Sussex. 2nd ed. Simpkin, Marshall, Hamilton, Kent and Co. Ltd., London. BENTHAM, G. & Hooker, J. D. (1954). Handbook of the British flora. 7th ed., revised by Rendle, A.B. Reeve & Co. Ltd., Kent. BRANWELL, A. E. (1872). Phyteuma spicatum — short notes and queries. Journal of botany, 10: 307-308. CLAPHAM, A. R., TuTix. T. G. & Moore, D. M. (1987). Flora of the British Isles. 3rd ed. Cambridge University Press, Cambridge. DAMBOLDT, J. (1976). Phyteuma (L.), in TuTIN, T .G. ef al., eds. Flora Europaea 4: 96. Cambridge University Press, Cambridge. Druce, G. C. (1932). The Comital Flora of the British Isles. T. Buncle & Co., Arbroath. ELLENBERG, H. (1988). Vegetation Ecology of Central Europe. Cambridge University Press, Cambridge. Etuis, R. G. (1986). Tetrads. B.S.B.I. news. 43: 9. FitzGERALp, R. (1988). Phyteuma spicatum in Southeast Region Rare Plants Project 1985/1987. English Nature, Lewes., unpublished. GERARDE, J. (1597). The Herball or Generall Historie of Plantes. John Norton, London. GERARDE, J. (1633). The Herball or Generall Historie of Plantes, Very Much Enlarged and Ammended, by Johnson, T. Norton and Whitakers, London. Gricson, G. (1958). The Englishman’s Flora. Hart-Davis, MacGibbon, London. HALL, P. C. (1980). Sussex plant atlas. Borough of Brighton, Booth Museum of Natural History, Brighton. JENNER, E. (1845). A Flora of Tunbridge Wells. Tunbridge Wells. Joint NATURE CONSERVATION COMMITTEE (1996). Third Quinquennial Review of Schedules 5 and 8 of the Wildlife and Countryside Act, 1981. Report and Recommendations from the Joint Nature Conservation Committee. June 1996. Joint Nature Conservation Committee, Peterborough. PARKINSON, J. (1640). Theatrum Botanicum. T. Cotes, London. PERRING F. H. & FarreLt, L. (1983). British red data books 1: vascular plants. 2nd ed. Royal Society For Nature Conservation, Lincoln. Ropwe LL, J. ed. (1991). British plant communities. Volume 1: woodlands and scrub. Cambridge University Press, Cambridge. RopweLL, J. ed. (1992). British plant communities. Volume 3: grasslands and montane communities. Cambridge University Press, Cambridge. Stace, C. A. (1997). New Flora of The British Isles, 2nd ed. Cambridge University Press, Cambridge. WHEELER, B. R. (1997). Aspects of the ecology and conservation of the rare plant species Phyteuma spicatum L. (Campanulaceae) in the British Isles. Unpublished. DPhil thesis, University of Sussex. Wiccinton, M. J. W. (1999). British red data books 1: vascular plants, 3rd ed. Joint Nature Conservation Committee, Peterborough. Wi.tems, J. H. (1980). De witte rapunzel in het Leudalgebied. Rondom het Leudal, 20: 16-18. Wo Ley-Dop, A. H. (1937). Flora of Sussex. Kenneth Saville, Hastings. (Accepted November 1998) PHYTEUMA SPICATUM IN BRITAIN a93 APPENDIX A list of records discovered for P. spicatum, compiled from the sources described above is given. 10-km square and, where possible, standard B.S.B.I. tetrad reference (Ellis 1986), is given for each site. Only those records of P. spicatum growing wild are included and consequently all records are from v.c. 14. Only first and last (most recent) records are given, except where interim records provide valuable information, for example on plant numbers. The source of the record is given either as a publication reference or as a herbarium or report reference. Records courtesy of Paul Harmes, the B.S.B.I. Recorder for v.c. 14, are indicated by BSBI. Records courtesy of English Nature, Sussex and Surrey Team are indicated by EN. Records are arranged numerically by tetrad reference. The symbol < before a date indicates that the record is earlier than the date given, but that its exact date is not known. In such cases the date of the publication or document in which the record was located is supplied instead. The symbol * against a record indicates a currently extant (June 1996) population of P. spicatum at the site. EAST SUSSEX, V.C. 14, Firle (TQ/4.0), 1968, (BSBI); seems unlikely, no records prior to or after this one. Near Glynde (TQ/4.0), (K). 1934, Pickard (BM); specimens of seedlings only in Herb. Musei Brittanica — seedling specimens difficult to confirm as P. spicatum; also G. C. Druce (Wolley-Dod 1937). Ringmer/Isfield, Plashett Wood, along ride (TQ/4.1M), 1967, A. W. Jones; searched by David Lang and Paul Harmes in 1993, but no plants found. Maresfield, a casual at (TQ/4.2), Rev. A. C. D. Ryder (Wolley-Dod 1937). Uckfield, Hempstead Wood (TQ/4.2V), <1900, J. H. A. Jenner (Wolley-Dod 1937). Uckfield (TQ/4.2), 1895, Druce (K). Arlington, Park Wood Farm (TQ5.0P), E. Bray (Wolley-Dod 1937). Near Hailsham (TQ/5.0), 1877, J. H. A. Jenner (BM); 1906, FE. Bray (K); probably either Tilehurst, Bramble Grove or one of the Wilmington Forest woods. Near Berwick (TQ/5.0), 1927, Miss Knox (BSBI); possibly a wood which has disappeared, or perhaps Plackett Coppice, in Wilmington Forest, the nearest known record. Upper Dicker, Bramble Grove (TQ/5.0U), 1906, E. Bray (BM); “There is a good bit of it...”, 1919, E. Bray in a letter to A. J. Wilmott (BM); Bramble Grove Wood searched by FitzGerald in 1987, no plants found, “uncoppiced for many years and very dark’, EN. Upper Dicker, Tilehurst Wood (TQ/5.0U), 1906, E. Bray (BM); 1919, E. Bray (BM); “access not gained”, FitzGerald (1987). Arlington, woodland (TQ/5.0]), 1885, J. H. A. Jenner (LIV); no exact location but could be referring to the nearby Wilmington Forest. Copse near Michelham Priory, between Hailsham and Abbots Wood (TQ/5.0P or U?), 1933, J. E. Lousley (RNG); could perhaps be Bramble Grove or Milton Hide. *Abbots Wood and Wilmington Forest (TQ/5.0N, T and U); there are numerous records for Wilmington Forest; most are Abbots Wood but many are Cane Heath, Nate Wood, Plackett Coppice, Milton Hide or Wilmington Wood -— it is often impossible from the herbaria specimens to determine the precise location; first and last records for the Forest as a whole are given here; 1873, F. C. S. Roper (BTN). (also 1875 “dist. abundant’); felled portion of Abbots Wood near Milton Hide 1942, P. W. Richards (EN); “roadside show” along edge of Abbots Wood near Old Oak Inn, 1968, A. W. Jones; TQ/5.0N — ten plants on edge of woodland path in mixed deciduous area, TQ/5.0T — one plant on bank of main forest ride, TQ/5.0T — eleven plants in coniferous woodland edge by path, TQ/5.0T — two plants in overgrown dark woodland on bank of ditch, TQ/5.0T — five plants in a large clearing in the centre of Abbots Wood, and TQ/5.0N — one plant at the side of the road at the car park entrance, 1996, B. R. Wheeler. Upper Dicker, Mill Wood, woodland ride (TQ/5.0P), 1954, D. Philcox (BSBI). Nate Wood (part of Wilmington Forest) (TQ/5.0T), 1962, P. Cockburn (BSBI); woodland by stream, 23 - plants in small patches, 1980, R. Minor, (BSB). Michelham Pnory (TQ/5.0P, U) (Amold 1907); “Originally perhaps an escape from Warbleton or Michelham Pnories...” (Wolley-Dod 1937); “It is still grown in the apothecaries’ garden’, and an exhibit in the museum claims “it was brought to Michelham by the canons from its native France’, 1986, R. FitzGerald (EN); the plant was not to be found in the apothecaries’ garden in 1996 when a colleague of the authors visited, 1995. 394 B. R. WHEELER AND M. J. HUTCHINGS Michelham Priory Wood (TQ/5.0U), 1924, Foggitt (CGE); most likely to be Bramble Grove or perhaps Mill Wood. Laughton Woods, Laughton. (TQ/5.1B), <1900, J. H. A. Jenner, (Wolley-Dod 1937). Bentley Wood, between Halland and East Hoathly, (TQ/5.1D), 1895, E. H. Farr (LIV); this record could refer to Bentley Wood at TQ/5.1D or at TQ/4.1Y; Druce’s record for The Nursery, west of East Hoathly is for the same wood, Bentley Wood was named Nursery Wood on early 19th century maps; “NW part of wood, past S. of stream at about 51/506169, not very far east of public footpath.”, C. A. Lister, pers. comm. to F. Rose, 1959, (EN); site searched by FitzGerald in 1987, no plants found, “very overgrown’, (EN). Wood near East Hoathly (TQ/5.1D, I, C or H?), 1896, 7. Hilton (BM); this could be any of a number of woods in the area but may be Bentley Wood since it has confirmed records. Bolt Wood, Chiddingly. (TQ/5.1L), 1930, M. Abbott-Anderson (LIV). Waldron (TQ/5.1P?), 1835 (Arnold 1907); “This is confined to a few square miles of the county and is not common there only a few hundred plants probably”, 1892, Farr (NMW). Little London (TQ/5.1U), H. S. Burder (Wolley-Dod 1937). Heathfield, woods at Stillyans (TQ/5.1Z), 1869, Dr Hogg (BM); Stillyans is the farm just below the wood in which the following records also occur; *Maynards Green (TQ/5.1Z); “Copse E. of Maynards Green’, 1956 K. E. Bull (BSBI); “Plentiful in narrow bit projecting from main wood’, 1957, K. E. Bull to F. Rose (EN), and “There were not so many plants as usual, perhaps a dozen to 20 scattered flowering plants”, 1959, K. E. Bull to F. Rose (EN); “Both patches N of the stream (two plants and 69 plants) in tree-fall clearings, plants S (seven plants) of the stream in deep shade and more scattered’, 1986, N. Stewart and R. FitzGerald (EN); Approximately 20 plants still to N of stream in main wood but Bull’s patch in the copse was not found, nor those S. of the stream, 1996, B. R. Wheeler, 1996. Maynards Green to Springdale Farm (TQ/5.1Z), 1965, R. A. Boniface (NMW); “This area is private gardens and coppices. Access not gained. Stream looks overgrown and dark”, 1987, FitzGerald (EN). New Sapperton Farm, Vines Cross/Maynards Green, twelve plants, (TQ/5.1Z), 1986, N. Stewart (BSBI); site not searched since, so possibly extant. Sapperton Manor Farm, Vines Cross/Maynards Green, stream bank (TQ/5.1Z), 1968, P. C. Hall (BSBI); this area has not been searched since. Knight’s Farm / Hadlow Down, on the estate of the late Mr Day, at Hudlow (Mayfield).....near the hedge of a hop-garden (TQ/5.2), Rev. R. Price, (Branwell 1872; Amold 1907); on Hadlow Down, nr Mayfield and Knight’s Farm, Mayfield, nr Cross-in-Hand, 1826, herb. Borrer (K); Hadlow Down, 1824, Rev. R. Price (herb. Borrer) (BM); Knight’s Farm, Mayfield, a mile from Cross-in-Hand, 1829, (Wolley-Dod 1937); Wolley-Dod says “these two are the same station, and though formerly in Mayfield Parish, it is fully four miles from that village’; ‘““The plant was growing in the wood and lane near the field, formerly a hop-garden, last July, A. E. Branwell (Branwell 1872). Mayfield (TQ/5.2), 1887, Rimington (CGE); “Between Mayfield and Broadoak”, T. Atkinson, (Wolley- Dod 1937). Heathfield, on the old iron-foundries (TQ/5.2), E. Head in Wolley-Dod’s notebook, (BTN); also in Notes on herb. Borrer. (BTN). Heathfield, streambanks south of (TQ/5.1/TQ/5.2) (Wolley-Dod 1937). Heathfield Golf Links (TQ/5.2), Mrs Morton. (Wolley-Dod 1937). Old Heathficld, corner of Mill Pond, (TQ/5.2?) no date but pre-1980 (BSBI). Hudlow (Hadlow Down?), “I have found it in hedgerows, scattered for miles, near Hudlow” (TQ5.2), A. E. Branwell (Branwell 1872). Framfield, Pound Lane, “both sides of lane going north from A272, perhaps 1/3 of distance from main road to fork of lane’, (TQ/5.2A), 1949, B. Welch, (BSBI) — this site refers incorrectly to the A272 — the road at those co-ordinates stems from the B2102 and joins the A272 further north. Framfield, between Pound and Bndgelands Farm (TQ/5.2A), 1974, P. B. Clarke (BSBI). Roadside verge, 1983, P. Donovan (BSBI); site searched in 1986 by FitzGerald, (EN), and in 1987 by E. J. Rich —no plants found. *Dower House Farm, Blackboys, woodland edge (TQ/5.2F), 1986, P. Donovan (EN); edge of ditch at woodland edge, 1996, D. Gunner. Possingworth, Waldron Down (TQ/5.2F or K), no date, J. Woods (BM); “Borders of Possingworth Wood”, Mrs Baines, (Wolley-Dod 1937). PHYTEUMA SPICATUM IN BRITAIN 395 *Tinkers Lane, Cross-in-Hand (TQ/5.2G); “‘literally 100’s of flowers..”, 1967, E. J. Ashdown (courtesy of T. G. C. Rich), 1970; steep roadside bank off lane called Tinkers Lane, 285 adult plants in 1996 — main concentration at eastern end of lane with smaller numbers spread along verge, 1996, B. R. Wheeler. Hole Farm, “In the middle of a wood on the Hole Farm in Buxted Parish...”, (TQ/5.2G), 1835, Bromfield (herb. Roper) (BTN); “In Hole Wood on border of field between that and Knight’s in great abundance.”’, 1941, E. Forster (BM). Loudwell/Little England Farm, Hadlow Down. (TQ/5.2L), — the woods are owned by Little England but often recorded under the nearby Loudwell Farm — “Stream at edge of small wood SE of Loudwell Farm.”, 1974, P. C. Hall (BSBI); 50 plants counted, 1980, E. J. Rich, (BSBI); eight plants found in “overgrown disused coppice’, 1986, N. Stewart and R. FitzGerald, (EN); area searched by B.R. Wheeler in 1996 but no plants found, area very dark. *Loudwell/Little England Farm, Hadlow Down. (TQ/5.2L), 16 plants found, “..some daylight is needed.”, 1986, N. Stewart and R. FitzGerald, (EN); 5 vegetative plants found in dark overgrown coppice on streambank at Western corner of Homegrove Wood, 1996, B. R. Wheeler. Mill Lane, Cross-in-Hand, roadside verge (TQ/5.2K), 1973, E. Norman (BSBI); not found in 1980 by R. Minor, (BSBI). *Malls Bank, Nursery Lane, Cross-in-Hand (TQ/5.2K), 1974, P. C. Hall, (BSBI); plants on a roadside verge where the verge borders Malls Bank Farm — seven plants divided between two small areas of the verge, B. R. Wheeler. This is a reduction from 16 plants in 1980, R. Minor, and 18 in 1987, R. Fitzgerald and E. Wood, (EN). Selwyns Wood (TQ/5.2K), 1943, F. Rose (NMW). A lane east of Hadlow Down (TQ/5.2L or M?), (Wolley-Dod (1937) — this record could be the Knight’s Farm record. Cross-in-Hand (TQ/5.2Q?), 1832, W. Christy (CGE); also 1892, E. H. Farr (BTN); 1924, Wolley-Dod (BM); these could be a number of extinct or extant sites in the Cross-in-Hand area. *Brown’s Lane, nr Cross-in-Hand, 51 plants, Plants on south bank of lane — not in adjoining field. (TQ/5.2Q), 1980, R. Minor (BSBI); >55 plants in 1986, (EN); only 4 plants by 1994, B. R. Wheeler; overgrown hedge which had been shading verge in recent years had been cut in 1996 — it is possible that the verge may recover. Tilesmore Wood (TQ/5.2Q or R), 1827, W. Borrer (Wolley-Dod 1937). Field borders east of Dunsley Wood (TQ/5.2Q or R?), 1827, W. Borrer, (Wolley-Dod 1937) — this is probably Dunly Wood. East of The Pheasantry, Heathfield. (TQ/5.2W), 1968, H. Hartwell (BSBI); site searched by FitzGerald in 1986, no plants found, (EN). Warbleton, in the woods at Warbleton in such abundance...(TQ/6.1), 1845, E. Jenner (Wolley-Dod 1937); only remaining Warbleton Parish sites are Rushlake Green and St Dunstan’s Farm. *St Dunstan’s Farm, three plants in woodland edge (TQ/6.1A), 1994, J. Simes; still extant, 1996, P. Harmes. Between Herstmonceux and Hellingly (TQ/6.1B?), 1929, Miss K. M. Morris (BM) — this record could be Park Wood. Park Wood, Hellingly, woodland ride (TQ/6.1B), 1880, F. C. S. Roper (BTN); “Eastern ride towards Carters Corner.”, 1961, D. P. Young (BSBI). *Rushlake Green, single plant by path in woodland (TQ6.1J), 1990, M. MacFarlane and G. Stevens (BSBI); 4 flower spikes, 1995, J. Simes; unconfirmed for 1996, but likely to be extant, well known site. Warbleton Priory (TQ/6.1P) (Amold 1907). “Wood near Parson’s Mill” (TQ/6.2), 1835, W. M. Borrer (CGE) — exact location unidentified. Street End, Broad Oak (TQ/6.2B), no name (LIV). 1917, W. Byrne (BM); 1925, Mr Payne (BSBI). N NH ® 6 ip i i) se paral agueaA .,t at jaunt te DOG, me late, Nea RA ot ; rhe _ He. 5 H ttl woese pas pie Min ie! ate eh nid me rg ri te ¢og Lit AL r ,= r by -. i dies y i Ae Pia ae aie aot Shh ay —a - fi Via Al “Wik nee a Ha tos AL. ae sate one; a iet, ora aa Tera ; ap Rony ore “2; A 4 Lett Let ok oes ie : Colpaey ei a rey ys slut a i . a ie ree 1s Aa / - i i a pe, Al Abs Pavey: mie sonia | rity Ye an d me pis « (ene a : ogt? : Pia ies . ar a min 2 i wht 6 3 Peel evden Thy varios eee ow ip inh 6 thy fe 7 Tae it 43 TES E 5 at yi a y rt cihegdonds Sa F ee a Ca se ey Le Nie Gere veh sits ee ace. 7? - r , - = ty AP coarse Te & 2 wee? Soe ae oe chi we RID w ee ih . | PLA Asan iw rf ‘yor bok Sai) dates ht RO ttl A = SS , ; ie 2a ! i | ies lit ors f Vi ar 4 7 ae vs ] Ling nie 4 ré ' = y - y \ Z i ; - ; 7 a 7 ale POE ve i i ; vice E hag — : z . . . i Tu?) isles hk velo. ‘3 vy a wt he ¥ aay ih ort “tit i : F io ey Mee Piatt ee rar 4 } ay & t ‘ i , = a Watsonia 22: 397—403 (1999) 397 Conservation of Britain’s biodiversity: Cyperus fuscus L. (Cyperaceae), Brown Galingale i CGT RICH Plantlife, 21 Elizabeth Street, London SW1W 9RP ABSTRACT Cyperus fuscus is a rare plant in Britain and probably always has been. It has been recorded in a total of 13 sites in England and two in Jersey, but is currently known from six sites and one site respectively (c. 50% decline). It is an annual of bare, seasonally exposed, nutrient-rich, base-rich mud on the edges of ponds and ditches. It is at its northern limit of distribution and is probably limited by climate. Population counts for extant sites are given for the period 1993-1996. The numbers of plants vary from year to year and site to site, with many observers noting it is most abundant in hot, drought years. Four sites have numbers below the minimum required to conserve all polymorphic genes with a frequency of 0-05 in the population. Most sites are grazed by stock, but in some, scrub has been removed to improve the habitat. Only one site is not protected. Keyworbs: Rare species, conservation. INTRODUCTION Cyperus fuscus L. (Cyperaceae), Brown Galingale, is a rare plant in Britain. In 1992, the wild-plant conservation charity Plantlife became concerned that it was amongst the most threatened plants in Britain. It was therefore included in their “Back from the brink” project, which aims to conserve critically endangered plant species through research and management work. About 20 rare plant species have been included in this project between 1992 and 1996, which represents a significant contribution to the conservation of biodiversity in Britain by the voluntary sector. The aim of this paper is to summarize the conservation work carried out on C. fuscus to 1996; full details can be found in Rich (1993a, 1993b, 1994, 1995) and Rich et al. (1996). Further details about the ‘““Back from the brink” project can be obtained from Plantilife. DISTRIBUTION DISTRIBUTION IN BRITAIN The distribution of C. fuscus is mapped in Fig. 1. It has been recorded from 13 native sites in England (some of which have or have had more than one population) and two in Jersey, in a total of eleven 10 km squares. The English sites are concentrated along parts of the valleys of the River Thames and River Avon with outlying sites in Somerset, Dorset and the Weald. It was also introduced to Fulham Common from Swiss material by A. H. Haworth inc. 1819 and was reported regularly until 1865 when the meadow was drained and built on (Gray 1871); this site is not discussed further. A record for Guernsey probably refers to Jersey (McClintock 1975). The dates of first and last records are summarised in Table 1. It has only been recorded in six sites in England and one site in Jersey since 1990 (c. 50% decline). The reasons for its decline are not always clear, but loss of ponds, drainage, gravel extraction, land reclamation or natural in-filling are possible reasons for loss. Most of the decline took place by the 1920s, and only one site has been lost in the last 50 years. *Address for correspondence: Department of Biodiversity and Systematic Biology, National Museum and Gallery of Wales, Cardiff CFl1 3NP 398 T. C. G. RICH Ficure 1. Distribution of Cyperus fuscus in the British Isles. ™@ 1990-1996; D1 pre—1990; + introduced. WORLD DISTRIBUTION Cyperus fuscus is widespread in Europe, adjacent parts of Africa and large parts of Asia (the eastern limits are insufficiently known; Hultén and Fries 1986). It is rare and declining 1n northern Europe (e.g. extinct in Sweden; Lindberg 1977) but is quite common in parts of central Europe. For instance it is quite frequent along the Rhine in Germany (Haeupler and Schonfelder 1989); in the Czech Republic, it 1s a characteristic plant of fish ponds in the south of the country (e.g. Husak 1953): It is rare in the eastern United States of America, occurnng westwards to Nebraska and South Dakota (Weedon and Stephens 1969). It was found in Canada in 1970 as an adventive at the edge of a pond in Ontario Province (Gillett 1971). CONSERVATION OF CYPERUS FUSCUS 399 TABLE 1. DATES OF FIRST AND LAST RECORDS OF SITES OF CYPERUS FUSCUS IN THE BRITISH ISLES. EXTANT SITES ARE ONLY LOCALISED TO COUNTY Site No. of First Last Reason for loss populations record __ record Berkshire 1. | C982 1996 - Berkshire 2. Pangbourne ] 1911 1911 Exact site not known, possibly reclaimed for agriculture Buckinghamshire 1. 2 1906 1996 - Buckinghamshire 2. Huntercombe l 1906 1906 __ Exact site not known, possibly reclaimed for agriculture Dorset 1. Bere Regis ] 1893 1893 — Site still suitable Dorset 2. 1-2 miles from Bere Regis ] 1893 1893 Exact site not known, possibly reclaimed for agriculture Dorset 3. Wimborne ] 1929 1929 Exact site not known Hampshire 1. ] 1934 1996 - Hampshire 2. 2 1983 1996, = Hampshire 3. Blashford ] 1893 1893 Probably lost to gravel extraction Middlesex 1. ] 1957 SSG Somerset |. 3+ 1899 1996 1 population still extant; others possibly lost through succes- sion to fen/swamp vegetation Surrey 1. Shalford ] 1846 1960s Seedbank possibly still present, pond rarely dries out Jersey 1. St Peter’s Marsh ] 1842 1842 Site drained Jersey 2. ] 1989 1996 - ECOLOGY LIFE CYCLE C. fuscus is an annual. It probably germinates from early summer onwards when the seedbank in the mud becomes exposed by seasonal drops in water level. Some germination has been noted as late as August at sites in Hampshire and Buckinghamshire. Plants were seen in flower in early June in the Czech Republic (pers. obs., 1993), but in Britain it flowers later, from about July to September. Plants are wind-pollinated; the anthers are tiny and yellow, and are exserted after the styles. Plants vary in size markedly. Plants collected at Somerset | (see Table 1) last century were often lush and up to 30 cm tall, whilst those seen more recently there and elsewhere have often been only a few centimetres high, although flowering and fruiting freely. It can grow and complete its life cycle within four months, as shown by its persistence in ditches at Somerset 1 over a period of many years where the ditches were cleared out on a four monthly cycle (White 1912). It is possible to obtain two generations in cultivation in one year (R. S. Cropper, pers. comm., 1995). Plants fruit soon after flowering. Most fruits probably do not disperse outside the pond or ditch system, but fruits have been found in mud attached to birds (Salisbury 1970) and its distribution along river valleys suggests that fruit is also dispersed by water during floods. Evidence from conservation work suggests that it has a persistent seed bank (see below). HABITAT The plant occurs on the damp, open, seasonally exposed, muddy margins of small ponds and ditches. Most soils are nutrient-rich (e.g. Lousley 1976), and it may benefit from nutrients from wildfowl droppings. Soil samples measured from four sites ranged from pH 6-6—7-8. It will tolerate some salinity in Europe (pers. obs., 1993), but has not been recorded in saline habitats in Bnitain. It is often associated with annuals of disturbed mud such as Bidens cernua, B. tripartita, Gnaphalium uliginosum, Juncus bufonius, Persicaria maculosa, P. hydropiper, Ranunculus sceleratus, Rorippa nasturtium-aquaticum and R. palustris, and perennials such as Agrostis stolonifera, Alopecurus geniculatus, Glyceria fluitans and Mentha aquatica. The vegetation is 400 T. C. G. RICH TABLE 2. SUMMARY OF RECENT POPULATION COUNTS OF CYPERUS FUSCUS. NO PLANTS HAVE BEEN SEEN RECENTLY AT SURREY 1, BUT A SEED BANK MAY STILL BE PRESENT. * NOT SURVEYED SITE 1984-5 1993 1994 1995 1996 Berkshire | 1,000 250 35-40 30 60 Buckinghamshire 1] 0 41 188 96 7 Hampshire 1 1,500 1,000+ 10,000 25,000 5,000 Hampshire 2 36,500 c. 500 200 1,061 1,682 Middlesex 1 50 0 200 100 320 Somerset 1 2 5 33 10 ] Surrey | 0 0 0 0 0 Jersey 2 - 0 0 3 3,000+ Total 39,052 1,796+ 10,656 26,300 10,070+ usually open, but C. fuscus can sometimes be found under the canopy of taller swamp species, though it will not tolerate deep shade. It is sometimes found with other uncommon wetland species such as Hottonia palustris, Oenanthe aquatica, Persicaria minor and Rumex palustris, and in central Europe such communities (Cyperetalia fusci and Cypero Limoselletum) are regarded as botanical gems (Ellenberg 1988). CLIMATE Cyperus fuscus is at the northern limit of its distribution in Britain, and seed production is probably limited by climate indirectly through habitat conditions, and directly through effects on growth and reproduction. Many observers (e.g. Druce 1926) note that it is most abundant in hot, drought years, and this has been confirmed in recent years (e.g. Table 2). With a generally wet and cool climate in Britain compared to Europe, its wetland habitats are dependent on seasonal lowering of the water table through low rainfall and higher temperatures, thus exposing the mud to allow plants to germinate. Tutin (1953) noted that in cultivation at Leicester it required relatively high temperatures for germination, and set little or no seed in a cool summer. The high nutrient status of many of its sites may also enable rapid growth under suitably warm climatic conditions. It flowers from peak summer onwards in Britain and thus has a very short flowering season before the autumn rains begin and plants are flooded. Plants at Middlesex 1 have been observed to survive short periods of inundation by water, but not longer periods; plants collected in September 1994 after c. | week under water and transplanted to the Seed Bank at Wakchurst Place died and set no fruit. Cyperus fuscus appears likely to benefit from global warming if the climate becomes warmer and dricr, but probably not 1f it becomes warmer and wetter. POPULATION SIZES The population sizes at the seven extant sites in 1993-1996 are summarised in Table 2, with some earlier 1984-1985 data for comparison from surveys carried out for the Nature Conservancy Council (Everett 1987; L. Farrell, pers. comm., 1993). It was present in seven sites between 1993 and 1996, and a seed bank may be still present at Shalford Common. The number of sites present each year depends on the weather and on disturbance. All recorders find that it varies in abundance within sites from year to year. For instance Lousley (1976) noted that C. fuscus fluctuated in abundance from year to year at Shalford “from great abundance in years like 1949, when the pond was almost dry, to complete absence when the water is high or the pond is overgrown with tall vegetation”’. CONSERVATION SITE MANAGEMENT Most sites are usually subject to light disturbance which helps to maintain them in a generally suitable open condition, such as by cattle trampling (though this must not be excessive) (Table 3). CONSERVATION OF CYPERUS FUSCUS 401 TABLE 3. CONSERVATION STATUS AND CURRENT THREATS TO CYPERUS FUSCUS SITES. NNR = NATIONAL NATURE RESERVE. SSSI = SITE OF SPECIAL SCIENTIFIC INTEREST. SSI = SITE OF SPECIAL INTEREST (JERSEY) Site Status Management Threats Berkshire | SSSI = Cattle grazed. Clearance of willows in pond by English Nature/ National Trust. Buckinghamshire 1 None __ Heavily cattle-grazed common land. Possible changes in water table due to Maidenhead flood relief scheme; pollution from road run-off; significant increase or decrease in stocking levels. Hampshire | SSSI Cattle-grazed. Some ditch clearance Possibly pollution from road run-off; has been carried out to provide significant increase or decrease suitable water levels. in stocking levels. Hampshire 2 SSSI —_Cattle- and pony-grazed common Spread of Myriophyllum aquaticum; land. significant increase or decrease in stocking levels. Middlesex 1 SSSI Cattle-grazed common. Spread of Crassula helmsti and Clearance of willows by Myriophyllum aquaticum; Plantlife. pollution from road run-off; significant increase or decrease in stocking levels; trampling by fishermen; lack of management. Somerset | NNR __ Ditch in horse-grazed pasture. Inappropriate ditch maintenance; marked changes in water table. Surrey | None Oldcommon no longer grazed. Local residents require a fish and Margins disturbed in 1989 with- duck pond which 1s in conflict out success; pond dredged in with the Cyperus requirements. 1992 and some willows cleared, but pond has not dried out since. Jersey 2 SSI Ungrazed common. Reeds clearedin Changes to local water table; parts annually, and also mown. cess-pit effluent; lack of management. C. fuscus 1s not grazed by horses and cattle primarily due to its small size. The associated trampling may result in loss of some individuals, but these losses are probably compensated for by maintenance of short open vegctation which is suitable for other individuals. It has also benefitted from conservation work at Middlesex 1 and Berkshire 1 which had become overgrown with scrub. Clearance of tall dense willow scrub at Middlesex 1 in September 1993 resulted in the reappearance of the plants in 1994 (no plants had been observed at this site since 1989); further clearances were carried out in 1996. Clearance of willow scrub at Berkshire 1 in 1994 has produced a less spectacular response, but plants are recolonising newly exposed mud. Plants were found at Jersey 1 after clearance of reeds in an old pond. The timing of management work 1s critical. Clearly it should not be carried out when plants are growing, but at other times of year ponds are often too wet to work in safely. Experience has shown that disturbance late in the season produces good results the following year. Disturbance could also be carried out immediately the mud is exposed in early summer prior to germination. In Somerset 1, C. fuscus was reported to benefit from ditch clearance as late as June (White 1912). Calculations have shown that a minimum sample size of 172 plants is required to preserve all, or very nearly all, polymorphic genes with frequency over 0-05 in a population (Lawrence et al. 1995a, b). It is thus proposed that conservation management should aim to achieve at least 172 C. fuscus plants at each site each year. On this basis, three sites have populations consistently above the minimum sizes, and four below (Table 2). 402 iE GSRICH STATUTORY PROTECTION Table 3 summarises the protection and threats to each site. C. fuscus is protected under the Wildlife and Countryside Act 1981, which should prevent deliberate uprooting and collection. Five extant sites are protected as statutory Sites of Special Scientific Interest in England, though only one of these is specifically for C. fuscus; one site has no protection. Jersey 1 is a statutory Site of Special Interest. This species is still under threat in Britain. A Species Action Plan is currently being drawn up by Plantlife for English Nature. MONITORING AND RESEARCH It is essential that populations are monitored each year to determine the results of the conservation work, assess natural variation due to weather and to watch out for new threats to sites. Between 1993 and 1996 monitoring was carried out cost-effectively by simply counting plants, taking photographs and making observations on management with the help of volunteers. In the longer term, population sizes should be correlated against weather patterns and pond water levels to determine how close the links are between population peaks and good weather and vice versa. Research should also be carried out into seed set and germination under different environmental conditions (temperature, water-logging, etc.). ACKNOWLEDGMENTS I would like to thank all those who have helped with information and conservation work, especially John Alder, Clive Chatters, Robert Cropper, Jin Danihelka, Ruth Davis, Geoff Dawes, Lynne Farrell, Geoffrey Field, Mike Freeman, Peter Grainger, Paul Harmes, Nick Hinson, Jan Kirschner, Mark and Clare Kitchen, John Knight, Frances Le Sueur, Margaret Long, Martin Love, Peter Marren, Andy McVeigh, Steve Parker, David Pearman, John Ounsted, John Pinel, Ron Porley, David Rees-Jones, Sarah Richardson, Tony Robinson, Francis Rose, Alan Showler, Dave Simpson, Shaun Swarbrick, Joyce Smith, Alison Templeton, Janet Terry, Audrey Thorne, Peter Tinning, Elizabeth Young, Heather Winship and Martin Wigginton. I would like to thank Del Monte, English Nature, Greenhams, the Royal Botanic Garden Wakehurst Place, and Spelthorne Natural History Society for access permission and help, and English Nature for access to files. The work was funded by Plantlife and Spelthorne Borough Council. REFERENCES Druce, G. C. (1926). The flora of Buckinghamshire. T. Buncle & Co., Arbroath. ELLENBERG, H. (1988). Vegetation ecology of central Europe. 4th ed. Cambridge University Press, Cambridge. Everett, S. (1987). Rare plant survey of South Region. Confidential report to Nature Conservancy Council, Peterborough. Gray, J. E. (1871). Cyperus fuscus not a native. Journal of botany 1871: 148 Gittett, J. M. (1971). Cyperus fuscus L. new to Canada. Canadian field-naturalist 85: 190. HAEUPLER, H. & SCHONFELDER, P. (1989). Atlas der Farn- und Bluten-pflanzen der Bundesrepublik Deutschland. Ulmer, Stuttgart. HuLTEN, E. & Fries, M. (1986). Atlas of north European vascular plants, north of the tropic of cancer. Koeltz Scientific Books, Konigstein, Germany. Husak, S. (1953). The Lednice fishponds. Propagacmi Tvorba, Praha. LINDBERG, P. (1979). Cyperus fuscus in Sweden. Svensk Botanisk Tidskrift 71: 69-77. Lousey, J. E. (1976). Flora of Surrey. David and Charles, Newton Abbot. McCurntock, D. (1975). The wild flowers of Guernsey. Collins, London. Ricn, T. C. G. (1993a). The status of Brown Galingale (Cyperus fuscus L.) in Britain. Back from the brink project report no. 14. Plantlife, London. Ricu, T. C. G. (1993b). Clearance at Shortwood Common for Brown Galingale (Cyperus fuscus L.). Back from the brink project report no. 15. Plantlife, London. Ricn, T. C. G. (1994). The status of Brown Galingale (Cyperus fuscus L.) in Britain in 1994. October 1994. Back from the brink project report no. 36. Plantlife, London. Ric, T. C. G. (1995). The status of Brown Galingale (Cyperus fuscus L.) in Britain in 1995. October 1995. Back from the brink project report no. 63. Plantlife, London. CONSERVATION OF CYPERUS FUSCUS 403 Ricu, T. C. G., et al. (1996). The status of Brown Galingale (Cyperus fuscus L.) in Britain in 1996. Back from the brink project report no. 74. Plantlife, London. SALISBURY, E. J. (1970). The pioneer vegetation of exposed mud and its biological features. Proceedings of the Royal Society 259: 207-255. TuTin, T. G. (1953). Natural factors contributing to a change in our flora, in LousLey, J. E., ed. The changing flora of Britain, pp. 19-25. B.S.B.I., London. WEEDON, R. R. & STEPHENS, H. A. (1969). Cyperus fuscus in Nebraska and South Dakota. Rhodora 71: 433. Wut, J. W. (1912). The flora of Bristol. John Wright and Sons, Bristol. (Accepted April 1998) li ton arrad. 4 at VE eet Le a iS) GS iis) Ay CF | ee Oe a ee ecaies- aetradens tai wd mene y~ PTS ee elias mi woich oul foe ew ideo v2. esuvely 29 Vary Goaiie oS. @ ih he bes WOR o1) G2) pp ae! ve scline ieee. end . aT Pi ruin = Ye ? - . : a: - vs Watsonia 22: 405-411 (1999) 405 Conservation of Britain’s biodiversity: Salvia pratensis L. (Lamiaceae), Meadow Clary PEG RICH* Plantlife, 21 Elizabeth Street, London, SW1W 9RP C. R. LAMBRICK Picketts Heath, Ridgeway, Boars Hill, Oxford, OX] 5EZ and C. McNAB 40 Nine Acre Close, Charlbury, Chipping Norton, Oxon., OX7 3RB ABSTRACT A summary of conservation work carried out on the statutorily protected plant species Salvia pratensis L. (Lamiaceae) in Britain between 1994—1996 is given. It has been recorded in 138 hectads and 32 vice-counties, mainly in southern Britain. It 1s accepted as native or probably native in 32 hectads and nine vice-counties, and introduced elsewhere. It has been recorded in 17 native or probably native hectads since 1990, and in another 15 prior to 1990, most of the loss of native sites occurring before 1950. It is a robust, polycarpic, perennial which sets abundant viable seed. It occurs in neutral or calcicolous grassland and scrub on calcareous soils. Sites may be unmanaged, mown regularly or irregularly, or may be grazed by stock or rabbits. It grows and reproduces best in sites which are winter-grazed with plants protected during flowering. Conservation work carried out includes population monitoring, introduction of suitable grazing regimes, soil scarification, scrub clearance, cross pollination and seed bank collections. Keyworbs: population sizes, habitat management, distribution, ecology. INTRODUCTION Salvia pratensis L. (Lamiaceae), Meadow Clary, is a handsome, blue-flowered, perennial herb which is typically found in unimproved, calcareous or neutral pastures. It was first reported from Kent in 1696 (Druce 1932), and is now known from sites scattered across southern Britain with the majority of the populations in Oxfordshire. It has also been recorded elsewhere as an introduction. S. pratensis has probably always been rare in Britain. Following concern that it was declining, surveys were carried out by the then Nature Conservancy Council during 1986-1988 (e.g. Everett 1987). It was added to Schedule 8 of the Wildlife and Countryside Act 1981, as amended, in 1992, and is statutorily protected. In 1994, Plantlife and the Ashmolean Natural History Society of Oxfordshire began a Species Recovery Programme with partial funding from English Nature to draw up a conservation plan for the species based on the excellent, detailed review of its ecology and conservation by Scott (1989). The aim of this paper is to summarise the conservation work carried out on the species in Britain between 1994 and 1996. Full details (which have been up-dated here) are given in confidential Plantlife reports (Rich 1995, Rich & McNab 1996, Rich et al. 1997). A full Species Action Plan is being implemented in England by Plantlife with funding from English Nature (Rich et al. 1997), and in Wales by the Countryside Council for Wales (L. K. Wilkinson, pers. comm., 1996). Various *Address for correspondence: Department of Biodiversity and Systematic Biology, National Museum and Gallery of Wales, Cardiff, CF1 3NP 406 T. C. G. RICH, C. R. LAMBRICK AND C. McNAB studies are also being carried out in Holland where the plant is also locally endangered (Ouborg, Van Treuren & Van Damme 1991; Van Treuren et al. 1991, 1993; Ouborg & Van Treuren 1995). DISTRIBUTION About 450 records of S. pratensis were traced in the literature, major herbaria (BM, BRISTM, CGE, K, LIV, MNE, NMW and OXF), Biological Records Centre and correspondence with various botanists, statutory agencies and Wildlife Trusts. Only a few records have been rejected as the plant is easily identified and many records are supported by herbanum specimens. The distribution of all records is summarised by 10-km squares in Fig. 1. S. pratensis has been recorded in 138 hectads (10-km x10-km squares) and 32 vice-counties. The records pick out the main areas of chalk on the North and South Downs, the Chilterns, the Hampshire-Wiltshire plains and the limestone of Oxfordshire and Gloucestershire, reflecting its general requirement for calcareous substrates. Whilst there is little doubt that S. pratensis 1s native in some localities in Britain and introduced to others, its status at some sites will probably never be resolved. The records have been revised as far as possible, noting the opinions of authors of the local Floras and other information. In general, S. pratensis 1s accepted as native in long-recorded sites or areas on chalk or limestone, and as introduced elsewhere. The presence of a single plant at a site may indicate that it is introduced, especially off calcareous soils, but the characteristic way in which single plants appear on suitable soils close to other records could also indicate a long persistent, native seed bank. Squares where S. pratensis is accepted as native are also shown in Fig. 1; it is native in 21 hectads and possibly native in another eleven hectads. In terms of vice-counties, it is native in vcc. 8, 15, 16, 17, 23, 24 and 34, and probably native in vcc. 13 and 35. This interpretation of the native distribution differs significantly from that presented in the Adlas of the British flora (Perring & Walters 1990). There has been some debate about the status of the Monmouthshire population (e.g. Riddelsdell 1916). Kay & John (1995) investigated the population genetics and demographic ecology of 32 species of lowland grassland and related habitats in Wales including S. pratensis. Nine loci in six enzyme systems were analysed in S. pratensis, of which five were variable. The pattern of heterozygosity was variable and unexpected, suggesting that recombination might be restricted by a degree of structural heterozygosity. An unrooted phylogenetic tree indicated that there was a high degree of correlation between geographic and genetic inter-relationships, with very little differentiation within the sub-divided population but relatively large distances between populations (N.B. one locality in Kent is erroneously treated as two localities). There was some evidence of a cline of decreasing genetic variability from the more vanable populations in south-east England to the least variable population in Monmouthshire. This pattern indicated that the surviving genotypic composition of the Monmouthshire population could be regarded as a typical edge-of-range population, perhaps derived from the Cotswold plants. This suggests that it could indeed be native or introduced from the Cotswold populations; based on the information available to date it is accepted as native. Further work on the genetics in relation to European material may provide more information on the status at this and other sites. There appear to be several modes of introduction. S. pratensis has been specifically noted as introduced with foreign grain to many sites in Britain, especially during the period 1880-1930. It is locally common in Europe and was a regular seed contaminant of imported grass/clover permanent pasture seed mixtures but was rare in grain from annually-cultivated arable fields. Salvia spp. have also been widely grown as herbs or salves (though S. pratensis is not especially noted for its medicinal uses). The name “Clary” appears to be derived from “‘clear-eye” — a practice by which mucilage from wetted seeds was put into eyes to “cleanse them” (this is not recommended practice). It was also used as a gargle for sore throats and as a cleanser of the teeth. The plant is now also fairly widely available in garden centres and its seed is present in so-called “‘wild-flower” seed mixtures. Some introduced records may therefore be a result of escapes from cultivation. Sturt (1995) has recently drawn attention to the association of Salvia verbenaca L., Wild Clary, with churchyards where it was sown on graves in the belief that it conferred immortality. At least eight records of S. pratensis are associated with churches, and this association may also account for some introduced records. CONSERVATION OF SALVIA PRATENSIS 407 ay th” FigurE 1. Distribution of Salvia pratensis L. in Britain. @ Native 1990 onwards; UO) native, pre-1990; * possibly native, 1990 onwards; o possibly native, pre-1990; *« introduced, 1990 onwards; + introduced, pre-1990. POPULATION SIZES At a national scale S. pratensis has been recorded as native or probably native in 17 hectads since 1990, and in another 15 prior to 1990. It has been recorded as an introduction in nine hectads since 1990. The plant should now thus be regarded as a Nationally Scarce species (cf. Stewart et al. 1994). Most of the loss of native sites appears to have taken place prior to 1950, and there is little evidence of significant decline in recent years; it is hoped that its statutory protection coupled with conservation work summarised below will prevent further losses. Populations were counted at each site at least once between 1994 and 1996; the maximum counts are shown in Table 1. Populations were usually counted by walking around sites, but at one large 408 T. C. G. RICH, C. R. LAMBRICK AND C. McNAB TABLE 1. MAXIMUM POPULATION COUNTS FOR NATIVE SALVIA PRATENSIS L. SITES RECORDED AT LEAST ONCE BETWEEN 1994 AND 1996. SITES ARE ONLY LOCALISED TO COUNTIES Site Year Population Buckinghamshire 1 1996 ] Gloucestershire 1 1994 200 Gloucestershire 2 1994 3 Gloucestershire 3 1996 l Kent 1 1994 20 Kent 2 1996 ©, 13) Monmouthshire | 1994 3 Oxfordshire 1 1994 4000-5000 Oxfordshire 2 1994 408 Oxfordshire 3 1995 43 Oxfordshire 4 1994 8 Oxfordshire 5 1995 1270 Oxfordshire 6 1996 20 Oxfordshire 7 1995 19 Oxfordshire 8 1995 4 Oxfordshire 9 1995 358 Oxfordshire 10 1996 6 Oxfordshire 1 1 1996 D211 Oxfordshire 12 1996 8 Oxfordshire 13 ° 1996 20 Surrey 1 1996 1] Sussex | 1994 200 Sussex 2 1995 ] Wiltshire 1 1994 1 site a more detailed systematic grid was used. In general each discrete clump was assumed to be one plant, but some clumps may have consisted of more than one individual. Counts included seedlings, vegetative and flowering plants. The population sizes vary enormously from single plants to a few thousands. An unusual feature of S. pratensis is the survival of a single plant in one remote locality for long periods, perhaps for 30 years or more. There are about 7000 plants in 23 native sites (Table 1), and about 2000 introduced plants in Britain (Rich 1995). An analysis of change in Oxfordshire indicates that 1994 populations are markedly below those of 1986-1988 censused by the Nature Conservancy Council (Everett 1987), though no sites have been lost (C. R. Lambrick in Rich 1995). The same appears to be true for some of the other populations elsewhere, indicating longer term climatic conditions may be important in regulating population sizes. It is likely that the management of sites 1s also critical. Ouborg & Van Treuren (1995) examined the relationship between population size and fitness in S. pratensis in the Netherlands. In theory small populations would be expected to have lower fitness than large populations as higher levels of genetic drift and inbreeding would occur leading to areduced viability and fecundity of the population. To test the theory, they cultivated plants from two large and two small Netherlands populations under similar conditions, and measured various components of seed size, germination, growth and reproduction. Contrary to the predictions, they found no evidence for reduced fitness in small populations, nor was there any evidence in the small populations for genetic erosion of fitness. It is possible that the small populations were at an early stage of genetic erosion where the allozyme diversity was low but the quantitative genetic variation underlying the fitness traits had not (yet) been affected. Applying these results to British populations, it may mean that even the small ones are viable in genetic terms. The evidence at the moment suggests that environmental .or stochastic processes have more significant effects on populations than genetic processes. CONSERVATION OF SALVIA PRATENSIS 409 ECOLOGY The population biology and ecology of S. pratensis in Oxfordshire, and at a few sites in Gloucestershire and Hampshire was described in detail by Scott (1989), from which the account below has largely been taken, with additional data from other sites (Rich 1995; Rich & McNab 1996). S. pratensis is a long-lived, polycarpic, perennial. Plants are relatively compact in growth form even when growing at high density (up to 50 plants per m *) and may be distinguished from one another by distinctive morphological characteristics which include the sex, length and colour of flowers. The peak flowering period is from late May to early July; Scott (1989) found 32-94% of the plants flowered on unmanaged sites. It is gynodioecious and self-compatible, though insects are required for within-flower selfing. Populations contain 1-9% of male-sterile plants and their frequency increases under intense shade or after damage from herbicides. Male-sterile plants have aborted stamens and smaller flowers than hermaphrodites. Flowers were visited by five species of pollinating bumblebees, and by pollen beetles, robber bumblebees and two species of butterflies. Scott (1989) found that seed-set exceeded 90% on most sites, but was slightly reduced where all plants were robbed (84%) or when fewer pollinators were available in mid-August (83%). Late flowering male-sterile plants set 67-74% of seed suggesting that even hermaphrodites are substantially out-crossed. Seed took one month to ripen and was released from mid-July onwards. Laboratory experiments showed that 72% germinate immediately after wetting in daylight or darkness, though after-nipened seed is less viable. Seed can germinate after at least one year of dormancy in the soil seed bank. Seedlings successfully establish where disturbance and stress generate gaps in the sward, exposing bare, uncompacted mineral soil. Short swards maintained by rabbit grazing provide very favourable conditions for seedling establishment and it is in these areas that the density of seedlings and mature plants is highest. Suitable niches for seedling establishment include worm casts, old ant hills, small mammal runs and rabbit scrapes. Establishment is generally unsuccessful in tall undisturbed grassland where more than | cm of moss or leaf litter accumulates under vigorous Species such as Brachypodium pinnatum and scrub. Reproduction is also unsuccessful where plants are mown or intensively grazed by stock while flowering, as their inflorescences are destroyed (Scott 1989). Many populations occur on flat to steeply sloping sites facing south or west, which provide a warm, sunny microclimate. They generally grow on shallow, calcareous soils overlying oolitic limestone or chalk, but may also occur (and persist) on mesotrophic, freely-drained soils. Plants have deep roots which extend into the C horizon of the soil. S. pratensis occurs in hay meadows, grazed pastures, scrub and open woodland, and on banks, verges and tracks. It has been recorded in a range of communities of the National Vegetation Classification (Rodwell et al. 1991 et seg.). It mainly occurs in unimproved species-rich calcicolous grasslands (CG2 Festuca ovina — Avenula pratensis grassland, CG3 Bromus erectus grasslands, CG4 Brachypodium pinnatum grassland, CG5 Bromus erectus — Brachypodium pinnatum grassland) but surprisingly also occurs and persists in mesotrophic grasslands (MGI Arrhenatherum elatius coarse grassland, MG7 Lolium perenne leys). Scrub or woodland edge communities in which it occurs are usually W21 Crataegus monogyna — Hedera helix scrub. Sheep and cattle at a high stocking density graze off all inflorescences and defoliate plants. In Oxfordshire, plants grazed in late June fail to flower later in the year, but 13% of flowering plants cut during hay making in mid-June flowered again in mid-August (Scott 1989). It appears that roe and muntjac deer browse inflorescences while rabbits ignore S. pratensis entirely. On some sites significant damage was caused by sawfly larvae and small mammals to flowers and unripe seed respectively. The plant is also grazed by slugs and snails, which may cause significant mortality in damp situations. Dormant basal buds on the rootstock develop rapidly if the main axis is damaged by grazing or cutting, and as part of the seasonal growth-cycle from July onwards. 410 T. C. G. RICH, C. R. LAMBRICK AND C. McNAB CONSERVATION MANAGEMENT S. pratensis survives for long periods under various management regimes. Sites may be neglected or unmanaged, mown irregularly (c.g. road verges) or annually (e.g. hayfields), or grazed by rabbits or stock. In general, the plants seem to grow best and increase where the sites are winter-grazed with plants protected during flowering. Unfortunately this regime is not practicable for many sites such as road verges. Scott (1989) has drawn up a series of recommendations to maintain and encourage the expansion of populations in existing sites; disturbance is required to maintain an open grassland sward, prevent plant litter and moss accumulating, and control the spread of invasive species. For site-specific conditions, the following management regimes were recommended: 1. Unmanaged sites should have a grazing or cutting regime introduced. 2. Mown sites should be cut in late July (early August) and the cuttings raked off. Soil scarification may also be required. 3. Sheep grazed sites should be grazed until early May, and then again from late July (early August) onwards. Grazing with cattle may be possible all year provided the stocking density 1s not too high. Plants can be protected by small enclosures during the summer if necessary. Every 4—5 years, the sites can be summer grazed. These recommendations are being applied in various sites under the Species Action Plan and the results will be documented in due course. In addition, scrub clearance has been carried out at four shaded sites which has resulted in improved flowering performance. Soil disturbance has also been carried out at three sites but as yet has failed to produce more plants from the seed bank. At one site a single female plant has been cross-pollinated with another plant of known ongin from another site. Reinforcement of populations is being carned out at two sites with stochastically-threatened populations to increase the populations to this size. Seed has been collected under licence from English Nature from six sites and deposited at the Seed Bank at the Royal Botanic Gardens, Wakehurst Place; further collections are planned. ACKNOWLEDGMENTS The work has been funded by English Nature’s Species Recovery Programme, the Sir James Colyer-Fergusson Charitable Trust and Plantlife. We would like to thank the Rare Plants Group of the Ashmolean Natural History Society of Oxfordshire, the National Trust, John Alder, Penny Angold, Bob Brocklehirst, Jack Chapman, Helen Coyte, Reg Crossley, Jo Dunn, Sue Erskine, Trevor Evans, Lynne Farrell, Richard Fitter, Ro Fitzgerald, Pel Fursdon, Gerald Gardiner, Joan Garlick, Beatrice Gillham, Paul Harmes, Graham Hart, Katherine Hearn, Derek Hill, Nick Hinson, Carol Hora, George Hutchinson, Ann Hutchison, Quentin Kay, Mark and Clare Kitchen, John Knight, Mark Knight, Brian Laney, David Lankester, Colin Lee, Andy McVeigh, Roger Mitchell, Elizabeth Norman, Tom Olliver, Richard Palmer, Alison Paul, Roy Perry, Eric Philp, Jack Pile, Joyce Pitt, Ron Porley, Andrew Proudfoot, Graham Roberts, Andy Scott, Alan Showler, Brian Starkey, Roddy Stevens, Peter Sturgess, Joe Trinder, Cathy Warden, Lindi Wilkinson, Phil Williams and Derek Wise for their help. We also thank the keepers of the Natural History Museum (BM), Bristol City Museum (BRISTM), Cambridge (CGE), Kew (K), Liverpool Museum (LIV), Maidstone Museum (MNE), National Museum and Gallery of Wales (NMW) and Oxford (OXF) for access to herbaria and libraries. The map was plotted using DMAP. CONSERVATION OF SALVIA PRATENSIS 411 REFERENCES Druce, G. C. (1932). The comital flora of the British Isles. T. Buncle & Co., Arbroath. Everett, S. (1987). Rare plant survey of South Region. Confidential report to N.C.C., Peterborough. Kay, Q. O. N. & Joun, R. F. (1995). The conservation of scarce and declining plant species in lowland Wales: population genetics, demographic ecology and recommendations for future conservation in 32 species of grassland and related habitats. Countryside Council for Wales Science report no. 110. March 1995. Countryside Council for Wales, Bangor. OusorG, N. J. & VAN TREUREN, R. (1995). Variation in fitness-related characters among small and large populations of Salvia pratensis. Journal of ecology 83: 369-380. Ousora, N. J., VAN TREUREN, R. & VAN Damme, J. M. M. (1991). The significance of genetic erosion in the process of extinction. 2. Morphological variation and fitness components in populations of varying size of Salvia pratensis L. and Scabiosa columbaria L. Oecologia 86: 359-367. PERRING, F. H. & WALTERS, S. M., eds. (1990). Atlas of the British flora, 3rd ed. B.S.B.1., London. Ricu, T. C. G. (1995). The status of meadow clary (Salvia pratensis L.) in Britain in 1994. Back from the brink project report no. 44. Unpublished confidential report to Plantlife, London. Ricu, T. C. G., et al. (1997). Meadow clary (Salvia pratensis) in 1996. Back from the brink project report no. 76. Unpublished confidential report to Plantlife, London. Ricu, T. C. G. & McNas, C. (1996). Meadow clary (Salvia pratensis) in 1995. Back from the brink project report no. 72. Unpublished confidential report to Plantlife, London. RIDDELSDELL, H. J. (1916). Salvia pratensis L. Report of the Botanical & Exchange Club of the British Isles 4: 426-427. RopweLL, J. S. et al., eds. (1991 et seq.). British plant communities. Cambridge University Press, Cambridge. Scott, A. (1989). The ecology and conservation of meadow clary (Salvia pratensis L.). M.Sc. thesis, University College, London. STEWART, A., PEARMAN, D. A. & Preston, C. D. (1994). Scarce plants in Britain. J.N.C.C., Peterborough. Sturt, N. (1995). Wild clary (Salvia verbenaca) in churchyards. B.S.B.1. news 68: 28-29. VAN TREUREN, R., BULSMA, R., OuBorG, N. J. & VAN DELDEN, W. (1993). The effects of population size and - plant density on outcrossing rates in locally endangered Salvia pratensis. Evolution 47: 1094-1104. VAN TREUREN, R., BULSMA, R. VAN DELDEN, W. & Ousora, N. J. (1991). The significance of genetic erosion in the process of extinction. 1. Genetic differentiation in Salvia pratensis L. and Scabiosa columbaria L. in relation to population size. Heredity 66: 181-189. (Accepted June 1998) a ci eae oa ; t abu cn Watsonia 22: 413-416 (1999) 413 Rumex X akeroydii — a new Dock hybrid F. J. RUMSEY Dept of Botany, Natural History Museum, Cromwell Road, London, SW7 5BD ABSTRACT A new Rumex L. (Polygonaceae) hybrid is described and illustrated. A single robust plant of R. x akeroydii Rumsey hybr. nov. (R. palustris Sm. x R. cristatus DC.) was discovered growing with its parents on the Inner Thames Marshes SSSI, Rainham, S. Essex in 1991. Keyworpbs: hybridisation, Essex. INTRODUCTION While surveying the distribution of locally and nationally scarce taxa on the Inner Thames Marshes SSSI (Site of Special Scientific Interest) at Rainham Marshes, S. Essex (Grid reference TQ/5.8) in the summer of 1991, a striking Rumex hybrid was discovered at the head of a then dry drainage ditch. The area, previously used as rifle ranges, grazing marsh and silt lagoons, borders commercial transport and light engineering works. At the time of the visit in late August, the site was very dry and appeared somewhat derelict as a result of the apparent cessation of grazing and other management. Scarce ruderal species of open vegetation such as Puccinellia rupestris (With.) Fernald & Weath., previously present in cattle-poached areas by drainage dykes, were absent. The drainage dykes which cross the grassland did, however, still locally support healthy populations of the nationally scarce docks R. maritimus L. and R. palustris Sm. The areas bordering on Ferry Lane, which skirts the site on two sides, support a rich alien flora and have long been a hunting ground for those interested in such plants. A Heracleum sp., identified as H. mantegazzianum Sommier & Levier, but perhaps not that species, has been naturalized here since at least the 1920s and the alien dock species Rumex obovatus Danser and R. patientia L. (as subsp. patientia) have both been reported in this general area (Jermyn 1974). The most frequent (and perhaps only) member of the R. patientia/R. cristatus group present in 1991 was R. cristatus DC. (specimens confirmed by Dr J. R. Akeroyd). Whether this is the result of an earlier mis-identification, or more recent colonisation and perhaps ousting of R. patientia by R. cristatus is impossible to state in the absence of any earlier voucher specimens. R. cristatus 1s, in my experience, much the commoner on waste ground on the clay soils along the Essex side of the Thames estuary and would seem to still be increasing. R. cristatus was infrequent away from the roadside but scattered plants could be found extending onto the grassland area at Rainham previously used for the rifle ranges and grazing. One such plant, a rather depauperate example at the head of a drainage ditch, was growing with an unfamiliar, striking, robust plant recognized in the field as a hybrid due to its intermediate appearance and sterility. The base of the ditch supported an open vegetation dominated by a large stand of R. conglomeratus Murr. with scattered plants of R. palustris and R. maritimus occurring within 25 m, further down the same ditch. A single example of the hybrid R. conglomeratus X R. palustris (R. x wirtgenii Beck), conspicuous by virtue of its highly abortive fruits, and apparently unrecorded previously in S. Essex, also occurred in this ditch. The stature of the ditch-head hybrid, with shoots up to c. 1-5 m tall, and its close proximity to R. cristatus Clearly implicated that as one parent. In leaf and shape of the perianth segments the plant was Clearly intermediate between this taxon and R. palustris; the dentation of the margins of the perianth-segments (valves) ruled out R. conglomeratus and the teeth were coarser than would be expected had R. maritimus been involved. Furthermore, the plant did not show any sign of the distinctive golden coloration that R. maritimus assumes when fruiting. Somewhat surpmisingly, 414 F. J. RUMSEY given the ease with which Rumex species hybridise and the researches of many in the detection of such hybrids (e.g. Rechinger 1964; Lousley & Kent 1981), the cross between R. cristatus and R. palustris has apparently not hitherto been found. This is probably because their native ranges barely overlap (Jalas & Suominen 1979; Akeroyd 1993). Accordingly a name and illustrated description of this plant are given here. TAXONOMIC DESCRIPTION Rumex X akeroydii Rumsey, hybr. nov. (Rumex cristatus DC. x R. palustris Sm.) (Fig. 1) Hybrida inter Rumex cristatus DC. et R. palustris Sm., characteribus inter parentes variantes; planta robusta, valvae fructiferae 4-7 mm longa, acute et irregulariter denticulatae, dentibus angustibus ad summum 2 mm, seminibus praecipue sterilibus. Robust, erect perennial herb up to c. 1-5 m tall, the basal leaves narrowly ovate-lanceolate, cuneate. Inflorescence moderately dense, with erecto-patent to patent branches, leafy throughout. Valves cordate, suborbicular to triangular, 4-7 mm long, denticulate to dentate with spreading narrow teeth to c. 2 mm long. All valves with distinct tubercles, one larger, the other two subequal. Ho.oryeus:, Rainham Marsh, at head of drainage ditch, S. Essex, v.c. 18, TQ/520.813, putative parents close by, 27 August 1991, F. J. Rumsey (RNG); isotypus (BM). The specific name honours Dr John Akeroyd, who confirmed the identity of this hybrid, has done much to raise interest in this genus in Bnitain and Ireland, and like this taxon is a robust denizen of waste places. DISCUSSION This new hybrid is only likely to be confused with R. palustris x R. patientia (R. X peisonis Rech.), a hybrid as yet unreported from the Bnitish Isles. R. x akeroydii would be expected to show more irregularly dentate valve margins, but might only reliably be discriminated by cytogical studies, R. cristatus having 2n = 80, R. patientia 2n = 60 (Lousley & Kent 1981). The recognition of R. cristatus as specifically distinct from R. patientia has been questioned (Stace 1997). Akeroyd (1993) adopted a conservative approach and maintained it, with some reservations. In its consistently dentate valve margins, darker nutlet colour and leaf venation more nearly alt nght angles to the midrib, R. cristatus can always clearly be morphologically separated from the variable R. patientia. These characters and the cytological differences referred to above argue for their continued recognition at specific level. R. x akeroydii could occur anywhere in the sympatric range of the parent species (Fig. 2). R. palustris has a wide range across central and southern Europe, extending northwards to Denmark and southwards to central Greece. R. cristatus is restricted to Greece and the Aegean region, southern Albania, Cyprus and Sicily. The two only naturally overlap in the southern Balkans, where R. palustris is rather uncommon. The apparent lack of this obvious hybrid in the region may be the result of under-recording but probably reflects a genuine absence, the species being effectively isolated by habitat preferences. This is also largely true in Britain, where neither species is at all common but both may be locally abundant. In Britain R. palustris is almost restricted to wet nutrient-rich mud exposed in the late summer and autumn (Mountford 1994), whereas R. cristatus is a plant of dry ruderal waste places. In only a few localities will the species grow in close enough proximity for pollination and hybridisation to occur. The grazing marshes of the Thames estuary, with their cattle-disturbed wetlands, often in proximity to industrial sites, railways, roads, etc., provide suitable habitats for the formation and establishment of such Rumex hybrids. A range of very rare hybrids, such as R. crispus X R. maritimus (R. x fallacinus Hausskn.), R. conglomeratus X R. maritimus (R. x knafii Celak) and R. obtusifolius X R. cristatus (R. x lousleyii Kent) have been recorded from the Rainham, Pitsea-Bowers Gifford and Hadleigh Marshes further east (see Jermyn 1974, Stace 1975), to which can now be added R. x akeroydii. RUMEX X AKEROYDII 415 r +4 cones pit ir rn FiGurRE 1. Rumex X akeroydii. A. Fruit, i.e. perianth enclosing nutlet. B. Diagrammatic section through fruit to show relative size of tubercles. C. Basal leaf. D. Inflorescence. 416 F. J. RUMSEY “< Rumex =| cristatus = Rumex palustris Ficure 2. Distribution of R. palustris and R. cristatus (somewhat diagrammatic — based on Jalas & Suominen (1979)). The current status of the plant seen in 1991 is uncertain, as is the future of the site upon which it grew. Once foreseen as part of the route of the Channel Tunnel rail-link, then considered as the possible site of a theme park, the area is, in part, now the subject of a development proposal by Havering Council. | ACKNOWLEDGMENTS I am very grateful to John Akeroyd for determining the specimen and for his enthusiastic attempts to encourage my interest in this genus. REFERENCES AKEROYD. J. R. (1993). Rumex L., in Tutix, T. G. et al., eds. Flora Europaea 1. 2nd ed., pp. 99-107. Cambridge University Press, Cambridge. Jacas. J. & Suominen, J. eds. (1979). Atlas Florae Europaeae. Vol. 4. Polygonaceae. Committee for the mapping of the flora of Europe and Societas Biologica Fennica Vanamo, Helsinki. JERMYN, S. T. (1974). Flora of Essex. Essex Naturalists Trust Ltd., Colchester. LousLey, J. E. & Kent, D. H. (1981). Docks and Knotweeds of the British Isles. B. S. B. 1. Handbook no.3. Botanical Society of the British Isles, London. Movuntrorpb, J. O. (1994). Rumex palustris Smith, in STEWART, A., PEARMAN, D. A. & Preston, C. D., eds. Scarce Plants in Britain, p. 361. Joint Nature Conservation Committee, Peterborough. RECHINGER, K. H. (1964). Rumex L., in Tutin, T. G. et al., eds. Flora Europaea 1, pp. 82-89. Cambridge University Press, Cambridge. Stace, C. A., ed. (1975). Hybridization and the flora of the British Isles. Academic Press, London. Stace, C. A. (1997). New Flora of the British Isles, 2nd ed. Cambridge University Press, Cambridge. (Accepted November 1998) Watsonia 22: 417-419 (1999) A417 A new Epilobium hybrid from Scotland, E. pedunculare A. Cunn. x E. montanum L. D. R. MCKEAN Royal Botanic Garden, Inverleith Row, Edinburgh EH3 5LR ABSTRACT A new hybrid, Epilobium x kitcheneri McKean hybr. nov. (Onagraceae), between a widespread British species, Epilobium montanum L., and a much more local alien species E. pedunculare A. Cunn. is described from its one wild location in West Perthshire, v.c. 87. This new taxon more closely resembles the New Zealand parent but is generally larger in its parts and the shoots are ascending towards the apex instead of being prostrate as in E. pedunculare. Keyworps: New Zealand, Onagraceae INTRODUCTION One new hybrid Epilobium (Onagraceae) is described, which now brings the number of hybrids involving New Zealand species in Britain to six (Kitchener & McKean 1998). The others all have FE. brunnescens (Cockayne) Raven & Engelhorn as the New Zealand parent but this latest hybrid has the much less widespread FE. pedunculare A. Cunn. (E. linnaeoides Hook. f.) as one of the parents. This latter species 1s found on wet ground in open habitats mainly in W. Perth (v.c. 87), W. Galway (v.c. H16) and W. Mayo (v.c. H27); in England and Scotland it is a very local weed in gardens (Stace 1997). The other parent, E. montanum, 1s widespread throughout Britain and Ireland. The few well formed seeds of the hybrid (Fig. 1F) are similar in size and sculpturing to those of EF. montanum, with the papillae evenly distributed. The seeds of E. pedunculare are about 0-8 mm long, uniformly papillose on the convex side, and the ends are pointed, with the apex obtuse and the anterior end pale in colour and more acute, similar in shape to E. ciliatum Raf. seeds (cf. Stace 1997, Fig. 445 no. 4). The flat side is smooth except for a pale-coloured ridge between the two ends. This hybrid is fairly distinct from the hybrids with E. brunnescens because the leaves are much larger and the leaf margins are sharply toothed as in E. pedunculare. E. x kitcheneri McKean, hybr. nov. (E. pedunculare A. Cunn. x E. montanum L.) Hybrida inter E. pedunculare A. Cunn. et E. montanum L., characteribus inter parentes variantibus; planta semi-prostrata, effusa; stigmata lobis clavatis vel breviter lobatis ferentia; caulibus pilis crispatis brevibus uniformiter obtectis, folia glabra, margine ciliato, ovario pilis ad pedicellum subtentum extensis; florum colore pallide roseo, et seminibus praecipuc sterilibus sed interdum fertilibus tum plus quam 1 mm longis. A semi-prostrate straggling herb c. 20 cm long and 8 cm tall, rooting at the lower nodes and occasionally branching. Stem terete, generally clothed in dense short crisped hairs. Lower leaves Opposite, upper ones alternate, ovate with prominent teeth, 4-11 on each side, slightly bronze coloured on the underside, largest 2:5 x 1-8 cm, glabrous except margins ciliate, base more or less rounded, attenuate to petiole; petiole 3-5 mm long. Sepals linear-lanceolate, c. 0-35 mm; corolla light pink, c. 1-1 cm across with veined petals 7 mm long; 4 short stamens just reaching the stigma base and capable of shedding pollen there, 4 long stamens slightly overtopping stigma. Ovary c. 1-3 cm long with scattered short crisped hairs becoming denser on the pedicel. Style 0-4 mm, bearing clavate or shortly 4-lobed stigma. Capsule c. 1-7 cm, with scattered short crisped hairs, fruiting pedicel c. 2 cm long. Seeds mainly sterile, but with a few well developed ones c. 1-0—1-3 mm long. D. R. MCKEAN 418 vu Perea IT. ? 2 F397 JI VP ?2. SE \ Bea E “4 ys ; Os Lees ue = CSBEY anes SR a Va Ficure |. Epilobium x kitcheneri. A. plant; B. mid-stem; C. flower (half cut away); D. clavate style; E. 4-lobed style; F seed; G. transverse section of capsule. EPILOBIUM x KITCHENERI 419 Ho oryrus: Scotland, West Perth, v.c. 87, near Aberfoyle, off Duke’s Pass Road, dampish hillside track in oak wood, leading to abandoned quarry, grid reference NN/516.016, 18 August 1996. G. D. Kitchener, s.n., bar-code no. 00076122 (E). Paratypes (2) were cultivated from basal shoots of the wild plant and grown on to fruiting condition: 9 August 1997. G. D. Kitchener s.n., bar-code nos. 00076120 & 00076121 (E). The plant was growing in patches of Epilobium pedunculare (abundantly naturalised in the vicinity of Duke’s Pass) and both Epilobium montanum and E. obscurum Schreber grew nearby. The hybrid is named in honour of its finder, Mr Geoffrey Kitchener, an amateur botanist, who has been involved in the study of New Zealand willowherbs hybridising with British species. ACKNOWLEDGMENTS I would like to thank a local artist, Ms Louise Olley for her excellent illustration of the plant. I am also grateful to the finder, Geoffrey Kitchener who cultivated the plant and sent me much information about it. R. R. Mill gave advice and provided assistance with the Latin description and the Royal Botanic Garden (Sibbald) Trust provided a grant to pay for the art work. REFERENCES KITCHENER, G. D. & McKean, D. R. (1998). Hybrids of Epilobium brunnescens (Cockayne) Raven & Engelhorn (Onagraceae) and their occurrence in the British Isles. Watsonia 22: 49-60. Stace, C..A. (1997). New Flora of the British Isles. 2nd ed. Cambridge University Press, Cambridge. (Accepted March 1999) Watsonia 22: 421—432 (1999) 4? Notes ELYTRIGIA REPENS (L.) DESV. EX NEVSKI SUBSP. ARENOSA (SPENNER) A. LOVE (POACEAE) IN SOUTH-EAST YORKSHIRE (V.C. 61) The account of Elytrigia repens (L.) Desv. ex Nevski subsp. arenosa (Spenner) A. Love in north- western Europe (Trist 1995) prompted searches for this taxon on the north bank of the Humber estuary, including the Spurn peninsula, during the period 1995-1998. The following new records represent a significant extension to the known distnbution on mainland Bnitain:- Haverfield Quarry, TA/323.200; Welwick Bank, TA/335.193; Winsetts Bank, TA/380.177; Easington, near Firtholme Clough, TA/398.167; Spurn Bird Observatory, TA/420.148 and Spurn Warren, TA/408.117. Trist (1995) commented on the variable incidence of awns and both awned and awnless forms of this grass were recorded. At each location, E. repens subsp. arenosa occurs on consolidated sand of low salinity, cohabiting with at least four of the following: Erodium cicutarium (L.) L’Hér., Galium verum L., Potentilla reptans L., Ononis repens L., Allium vineale L., Carex arenaria L., Festuca rubra L., Agrostis stolonifera L., E. repens subsp. repens var. aristata (Doll) P. D. Sell, E. atherica (Link) Kerguélen ex Carreras Martinez, Ammophila arenaria (L.) Link and, on Spurn only, Potentilla anserina L., Plantago coronopus L., and Hippophaé rhamnoides L. These species indicate established and forming sand dune communities in the National Vegetation Classification SD7 through to SD10. DISCUSSION Knowledge of the distribution as presented by Trist (1995) suggests E. repens subsp. arenosa 1s rare in Britain and in north-western Europe. However, the ease with which these new records were taken suggests that E. repens subsp. arenosa may be yet under-recorded. Taxonomic rank as high as subspecies is considered doubtful by some, e.g. Stace (1991). However, as Trist (1995) shows, the taxon has had a “chequered history” ranging in rank from variety through subspecies to species. The Welwick Bank, Winsetts Bank and Easington locations, each of an area less than 400 m’, represent approximately 50% of the total SD7 to SD10 resource in v.c. 61. Each location is at nsk of damage as a result of proposed reconstruction of flood defence structures. Local rarity of habitat type alone may not be enough to protect these locations, even though they appear to be characterized, in part, by a low-rank taxon which can still be regarded as rare in Britain and north-western Europe. ACKNOWLEDGMENT With thanks to Dr T. A. Cope (Kew) for confirming the identification. REFERENCES Trist, P. J. O. (1995). Elytrigia repens (L.) Desv. ex Nevski subsp. arenosa (Spenner) A. Léve (Poaceae) in north-western Europe. Watsonia 20: 385-390. Stace, C. A. (1991). New Flora of the British Isles. Cambridge University Press, Cambridge. P. J. Cook Conservation Projects Officer, South Holderness Countryside Society 15 Park Avenue, Withernsea, East Yorkshire, HU19 2JX. E-mail: Coteds] @aol.com 422 NOTES THE POTENTIAL FOR SEED DISPERSAL BY SEA WATER IN COINCYA WRIGHTII (O. E. SCHULZ) STACE AND C. MONENSITS (L.) W. GREUTER & BURDET SUBSP. MONENSIS The two endemic Coincya taxa in Britain have restricted distnbutions (Rich 1991). C. wrightii (O. E. Schulz) Stace is confined to sea cliffs on the island of Lundy in the Bnistol Channel. C. monensis (L.) W. Greuter & Burdet subsp. monensis occurs On maritime sands around the east side of the Insh Sea from the Clyde to the Wirral and on the Isle of Man and has disjunct localities on Mull (extinct) and on the Gower and North Devon (the latter extinct). As both taxa have essentially maritime distributions, it might be expected that their seeds could be dispersed by the sea like other specialised maritime crucifers such as Cakile maritima Scop., Crambe maritima L. and Raphanus maritimus Sm. (e.g. Praeger 1913). An experiment was set up to investigate the potential for seed dispersal by sea water in the laboratory. In Coincya the fruits have two distinct parts which might have different dispersal capacities; the lower part 1s composed of twin loculi covered by dehiscent valves each with numerous seeds, and the upper part is an indehiscent beak with a few seeds. When the fruits are ripe, the valves dehisce to release the seeds in the loculi whilst the beaks remain intact until the infructescences break up. The experiments investigated the length of time for which individual seeds and beaks floated and the capacity for germination after periods of immersion in sea water and rain water. Only small numbers of seeds were available as both plants are rare. FLOATATION EXPERIMENT 100 seeds and 50 beaks of each taxon were placed in separate beakers of sea and rain water, and the number remaining floating recorded with time. The beakers were shaken at irregular intervals to simulate wave action. A small amount of domestic detergent (washing-up liquid) was added to some additional beakers to test if surface tension helped the seeds float. All beaks of both taxa floated for at least 12 hours, and some for up to four days, but most seeds sank immediately (Table 1). In both cases the proportion of beaks or seeds floating was higher in sea water than rain water as might be expected. Detergent caused all seeds and beaks to sink more rapidly (data not presented). GERMINATION EXPERIMENT Sub-samples of ten seeds were taken at intervals from beakers with sea water and fresh water, rinsed and placed on moist tissue paper in beakers to germinate in the light at room temperature. Controls were sown directly onto moist tissue without being immersed in water. Total germination (i.e. emergence of the radicle) was counted after 28 days. A few seeds in rain water went mouldy before germinating. The number of seeds germinating after immersion in rain and sea water for different periods of time is shown in Table 2. The first seeds germinated after six days when completely immersed in rain water. Seeds which had been immersed 1n sea water were generally much slower to germinate than seeds immersed in rain water. Germination was significantly lower in C. monensis than C. wrightii (ANOVA, all treatments lumped, d.f. = 1, p< 0.001). No seeds of either taxon germinated when immersed continuously in sea water for 28 days; dissection of the seeds in sea water showed that they had partially hydrated compared to dry seeds, whilst seeds in rain water appeared to have completely hydrated. There are no significant differences in germination after different periods of immersion in either sea or rain water for either taxon. CONCLUSIONS The floatation experiment shows that beaks float for longer periods than seeds in both taxa, and thus dispersal in sea water is more likely to occur by beaks than seeds. However, the length of time for which beaks float is very short, in general up to three tidal cycles (though one beak of C. monensis did float for four days in sea water), which may explain why both taxa have restricted distributions. NOTES 423 TABLE 1. NUMBER OF COINCYA BEAKS OR SEEDS FLOATING WITH TIME IN BEAKERS OF RAIN AND SEA WATER. Hours Coincya monensis Coincya wrightit Rain water Sea water Rain water Sea water Beaks Seeds Beaks Seeds Beaks Seeds Beaks Seeds n 50 100 50 100 50 100 50 100 0 50 7 50 24 50 17/ 50 34 1 50 3 50 8 50 14 50 26 2 50 l 50 5) 50 ] 50 19 3 50 ] 50 Dp 50 1 50 3} IW 50 1 50 0 50 1 50 1 24 0 0 20 0 48 0 45 0 36 0 0 11 0 7 0 10 0 48 0 0 5 0 0 0 0 0 MD. 0 0 3 0 0 0 0 0 96 0 0 1 0 0 0 0 0 120 0 0 0 0 0 0 0 0 TABLE 2. NUMBER OF COINCYA SEEDS GERMINATING AFTER IMMERSION IN SEA OR RAIN WATER FOR PERIODS OF TIME (N = 10). Time immersed Coincya monensis Coincya wrightii 0 days (control) D 10 Continuous sea water 0 0 Rain water Sea water Rain water Sea water 1 day 4 1 8 9 2 days 3 2 9 7 3 days 3 1 10 9 7 days 2 D fl 5 Seeds will not germinate in sea water, but seeds of both taxa retain viability even after seven days immersed in sea water, over twice the maximum floatation time. Presumably, as in Cakile (Hocking 1982), the high sodium chloride levels inhibit germination until seeds are washed ashore and leaching by rain lowers the salt content. ACKNOWLEDGMENT I would like to thank Ron Groom for supplying material of C. wrightii. REFERENCES Hockina, P. J. (1982). Salt and mineral nutrient levels in fruits of two strand line species, Cakile maritima and Arctotheca populifera with special reference to the effect of salt on the germination of Cakile. Annals of botany, new series 50: 335-343. PRAEGER, R. L. (1913). On the buoyancy of the seeds of some Britannic plants. Scientific Proceedings of the Royal Dublin Society 14: 13-62. Ricu, T. C. G. (1991). Crucifers of Great Britain and Ireland. B.S.B.1. Handbook No. 6. Botanical Society of the British Isles, London. TG) G: Rich 67 Heol Uchaf, Rhiwbina, Cardiff CF 14 6SR 424 NOTES SOUTHWARD RECOLONISATION BY MERTENSIA MARITIMA (L.) GRAY ON THE COAST OF NORTH-EASTERN SCOTLAND The distribution of Mertensia maritima, a beach plant growing in the strandline zone reached by highest tides, has fluctuated markedly since 1800 in northern Britain (Stewart 1994), with patterns quite different in the half-centuries 1800-1849, 1850-1899, 1900-1949 and 1950-1992. On the east coast between Fraserburgh and Berwick, the range of Mertensia contracted sharply after 1900, there being 28 occurrences in 10-km squares for the 1800-1899 period compared to just three occurrences for the 1950-1992 period (Stewart 1994). This author suggests that human recreation and shingle removal are the likely causes of this decline. Since about 1980, Mertensia maritima has been recolonising the east Aberdeenshire coast (v.c. 93) (Table 1, Fig. 1). Probably the species was totally extinct after 1950 between Fraserburgh and Aberdeen, since no records were made during the B.S.B.I. Maps Scheme (Perring & Walters 1962), but some stretches of this coast are remote and might not have been thoroughly examined. However, other shores would have been visited often, e.g. Craig Ewan by Peterhead Golf Course, or are in conservation areas (Sands of Forvie National Nature Reserve and Loch of Strathbeg R.S.P.B. Reserve) and got regularly recorded; for some of these shores there is no doubt that Mertensia was absent, then one to several plants colonised, and now numerous plants occur (Table 1). TABLE 1. POPULATION SIZE AND DATES OF RECORDING OF ELEVEN COLONIES OF MERTENSIA MARITIMA ON A 50 KM SECTION OF THE ABERDEENSHIRE AND KINCARDINESHIRE COAST Colony 1 km Distance (km) Year of Number of plants (plt) Recorder Square from Craig Ewan recording _and seedlings* (sdlg) Craig Ewan NK/12.48 0 1989 40 plts and sdlgs M. Innes 199] 35 plts and sdlgs M. Innes 1998 37 plts and sdlgs M. Innes Gadle Braes NK/13.46 l 1994 23 plts and sdlgs M. Innes 1998 50 plts and sdlgs M. Innes Sandford Bay NK/12.43 + 1985 5 plts D. Welch 1990 10 plts D. Welch 1998 3 pits D. Welch Furrah Head NK/13.43 5 1988 c.10 plts M. Innes 1998 c. 840 plts and sdlgs M. Innes Boddam NK/13.42 6 1989 7 pits D. Welch 1992 16 plts D. Welch 1998 absent D. Welch Whinnyfold NK/08.33 16 1992 few plts C. Millar 1998 absent M. Innes Perthudden NK/03.28 24 199] 3 plts + 25 sdlgs T. Dargic 1998 2 pits D. Welch Poor Man NK/03.27 24 1989 1 plt B. Davis 1992 14 plts + 83 sdlgs L. Farquhar 1998 19 plts + 4 sdlgs D. Welch Broadhaven NK/03.27 25 1992 9 plts + 32 sdlgs L. Farquar 1998 2 pits D. Welch Sanyne-Rockend NK/02.26 26 1992 73 sdlgs L. Farquar 1998 absent N. Harding Nigg Bay NO/96.04 48 1996 1 plt B. Ballinger 1998 1 plt D. Welch *seedlings were arbitrarily defined as having <10 leaves and being <10 cm’ in area. NOTES 425 Strahangles Point Ve FRASERBURGH — Craig Ewan PETERHEAD Oo Sands of Forvie NNR ABERDEEN O —— Nigg Bay 10 Km FicureE 1. Map of the coastal area of north-eastern Scotland showing main towns, the site of the long-standing colony of Mertensia maritima at Strahangles Point, and the section of coast between Craig Ewan and Nigg Bay being presently colonised. Evidence for a southward direction in the recolonisation is given by the first dates of observation in Table 1. At Craig Ewan the 40 plants present in 1989 indicate an initial colonisation c. 1980-1984; at Nigg Bay (v.c. 91) colonisation was almost certainly in 1996, judging from the small size of the single plant in that summer and the regular searching along this shore since 1992 made by botanists checking on a colony of Lathyrus japonicus. The source of the sea-borne seed initially producing these colonies could have been the large Orkney populations (Randall 1988; Stewart 1994) or the populations on the northern coast of Aberdeenshire and Banffshire (v.c. 94); the nearest of these populations extant after 1970 is at Strahangles Point, Aberdeeenshire (NJ/88.64), where 101 established plants and 66 seedlings were counted in 1987 (John Edelsten, pers. comm.). Differences in trends between sites over the study period are probably related to the coarseness of their substrate material. The colony at Furrah Head, which has increased spectacularly, occupies rocks and stones that have slumped to the beach from the boulder clay slope above, and at Poor Man the plants grow among large pebbles (mean diameter of c. 20 cm) but rooted in gravel. In contrast, sandy beaches have either not been colonised, e.g. the 22 km of coast between Rockend and Nigg Bay, or their colonies have remained tiny, e.g. at Sandford Bay. Here the three plants grow at the top of the beach, very close to a dense stand of Leymus arenarius which perhaps restricts colonisation; these plants may also suffer from being covered by sand during storms. 426 NOTES The few seedlings observed at the Forvie sites in NK/0.2 in 1998 suggest that establishment is difficult and variable between years. At Poor Man the largest plant had a diameter of 1 m in 1998, and the 18 other adult plants averaged about 50 cm diameter, so considerable quantities of seed should have been produced in the previous few years. Perhaps there was more germination in 1998 than was apparent at the September monitoring visit, and the seedlings failed to survive. Explanations for the observed distribution changes are necessarily speculative in the absence of monitoring on plant growth, longevity, fruiting and dispersion. With the species classified as scarce (Stewart 1994) and a total British population estimate of only c. 11,000 plants and seedlings in the 1980s (Randall 1988; Farrell 1989), determination of the controlling factors 1s very desirable. Low winter temperatures to stimulate seed germination, and relatively low summer temperatures to avoid drought for juvenile plants, have been put forward as controls on the broad range of Mertensia (Randall 1988). At the local scale other factors may be important. For the Nigg Bay colony, so strong in the nineteenth century that 20 separate specimens are known in British herbaria, Trail (1923) considered the cause of extinction was removal of shingle to make concrete for the south breakwater adjoining Aberdeen beach. Randall (1988), from visits to a wide range of colonies, considered that burial by sand in summer, grazing by sheep and rabbits, and trampling by humans could all severely deplete populations. But, with Mertensia now increasing at some sites with quite heavy public recreation pressures, it seems that even more factors may be affecting populations. We suggest that variable fruiting in response to climatic trends and shifts in the direction of sea currents also merit investigation as well as the factors advanced by previous workers. ACKNOWLEDGMENTS We thank Brian Ballinger, Tom Dargie, Bob Davis, John Edelsten, Lynne Farquhar, Nigel Harding and Colin Millar for allowing use of their unpublished records, and Lynne Farrell for commenting on a draft. REFERENCES FARRELL, L. (1989). Mertensia maritima (L.) Gray — current status in Britain. B.S.B./. news $1: 9-11. PERRING, F. H. & WacteRS, S. M., eds. (1962). Ailas of the British flora. Thomas Nelson & Sons, London. RANDALL, R. E. (1988). A field survey of Mertensia maritima (L.) Gray, Oyster plant, in Britain during 1986 and 1987. N.C.C. Contract Survey No. 20. Stewart, N. F. (1994). Mertensia maritima (L.) Gray. in STEWART, A., PEARMAN, D. A. & PREsTON, C. D., eds. Scarce plants in Britain, p. 267. Joint Nature Conservation Committee, Peterborough. TRAIL, J. W. H. (1923). Flora of the city parish of Aberdeen. Aberdeen University Press, Aberdeen. D. WELCH East Fernbank, Woodside Road, Banchory, Kincardineshire. AB31 5XL M. INNES 106A Queen Street, Peterhead, Aberdeenshire. AB42 6TY ERICA CILIARIS L. (ERICACEAE) DISCOVERED IN THE BLACKDOWN HILLS, ON THE SOMERSET-DEVON BORDER (V.C. 3) Erica ciliaris L. is a European near-endemic, with a world distribution extending from the north-westem tip of Morocco northwards through Portugal, western Spain and France, and reaching its northern limit in southern England, with one site in western Ireland. It is a member of the Oceanic Southern-temperate element of the British and Insh flora (Preston & Hill 1997). In Britain it has a curiously disjunct distribution, being almost entirely restricted as a native to the Purbeck area of Dorset (v.c. 9) and Commwall (v.cc. 1 & 2). Isolated records - usually of just a few plants — from Dartmoor, the New Forest and Anglesey are generally presumed, or known as in the case of Dartmoor, to have been introductions (Ivimey-Cook 1984; Rose et al. 1996; L. Spalton, NOTES 427 pers. comm.). In a few locations in the New Forest it is thought to be native and is extending its range naturally into suitable habitat in this area from its Dorset stronghold (Chapman & Rose 1994; Brewis et al. 1996). On 27th August 1998, during a routine visit to a Site of Special Scientific Interest (S.S.S.I.) in the Somerset part of the Blackdown Hills, the writer noticed a strange-coloured heather looking like E. ciliaris. A careful search revealed that EF. ciliaris, along with Erica tetralix L., was the commonest dwarf-shrub species over about 0-5 ha of wet heath and mire. The colony was clearly a long- established one, many plants being 0-4—0-6 m tall with shoots probably 15—20 years old. The total population of EF. ciliaris was difficult to estimate, but was thought to be in the order of 1,000—10,000 plants. The following day the author returned to the site with Paul Green and Jan Green, B.S.B.I. recorders for v.cc. 5 and 6 respectively, who confirmed that it was indeed E. ciliaris. Many plants of the hybrid between E. ciliaris and E. tetralix (E. X watsonii Benth.) were noted, with considerable variation in flower colour and foliage. Specimens of both E. ciliaris and E. x watsonii were sent to D. McClintock, who confirmed the identifications. Is it possible that F. ciliaris is native at this site, representing a geographical “missing link” between its strongholds in Dorset and Cornwall? The site is at an altitude of 210 m A.O.D., which makes it higher and further north than any other “native” English site. There is no evidence of E. ciliaris having been introduced — indeed, to those of us who have seen it there, the plant has every appearance of being native, occurring in an area of high-quality mire vegetation, with Molinia caerulea (L.) Moench., E. tetralix, Eriophorum angustifolium Honck., Succisa pratensis Moench., Narthecium ossifragum (L.) Hudson and Sphagnum spp. as common associates. In phytosociological terms the vegetation is mainly Narthecium ossifragum - Sphagnum papillosum valley mire (M21) with Potamogeton polygonifolius - Hypericum elodes soakways (M29), grading into Molinia caerulea - Cirsium dissectum fen-meadow (M24) on slightly dner ground (Rodwell 1991). This appears to be similar to the kinds of vegetation in which E. ciliaris occurs in Dorset (Rose et al. 1996). It is surprising that such a large and evidently long-established population of E. ciliaris could have been overlooked for so long. However, there may be several good reasons for this. Firstly, the general area is seldom visited by natural historians as it 1s not obviously en route to any of the region’s well-known botanical hunting grounds. Secondly, the site has no public access, and is not visible from any public nght of way. Thirdly, even supposing one had the good fortune to visit the site, the difficult terrain and tussocky nature of the vegetation would cause many fieldworkers to avoid the area in which E. ciliaris grows. Even so, it is extraordinary that the plant has been missed until now, given that over the last 15 years - during which time it must have been present - the site has received several visits from experienced field botanists, including Nature Conservancy Council and English Nature staff. On more than one occasion the area was visited in late August-early September, at a time when E. ciliaris should have been in flower! If, as suspected, its occurrence on this S.S.S.I. 1s a truly native one, there is a possibility that E. ciliaris will be recognised elsewhere in this region. There is other suitable-looking habitat in the Blackdown Hills, in both S. Devon (v.c. 3) and S. Somerset (v.c. 5), as well as on the East Devon Pebblebed Heaths, near Budleigh Salterton (v.c. 3). The newly discovered site for FE. ciliaris lies close to the Somerset-Devon border, in a part of Somerset that actually falls within v.c. 3 (S. Devon). The site is on privately owned farmland, and the owners have requested that details of its exact location should not be published. Anyone wishing to visit the site should contact the wmniter at the address below. ACKNOWLEDGMENTS I thank the owners for permission to visit the site and for their help in producing this note. Thanks also to Jan Green and Paul Green for their assistance with the field survey, to David McClintock for sO promptly determining the specimens sent to him, to Laurie Spalton (B.S.B.I. Recorder for v.c. 3) for comments on the final draft and to Simon Leach for extensive and very helpful comments on an early draft. 428 NOTES REFERENCES Brewis, A., BOWMAN, R. P. & Rose, F. (1996). The Flora of Hampshire. Harley Books, Colchester, in association with the Hampshire & Isle of Wight Wildlife Trust. CHAPMAN, S. B. & Rose, R. J. (1994). Changes in the distribution of Erica ciliaris L. & E. x watsonii Berth. in Dorset, 1963-1987. Watsonia: 20: 89-95. IvimEyY-Cook, R. B. (1984). Atlas of the Devon flora. The Devonshire Association for the Advancement of Science, Litereature and Art, Exeter. Preston, C. D. & Hitt, M. O. (1997). The geographical relationships of British and Irish vascular plants. Botanical journal of the Linnean Society, 124: 1—120. Ropwe LL, J. S., ed. (1991). British plant communities, Volume 2: mires and heaths. Cambridge University Press, Cambridge. Rose, R. J., BANNISTER, P. & CHapMan, S. B. (1996). Biological Flora of the British Isles: Erica ciliaris L. Journal of ecology, 84: 617-628. M. J. EDGINGTON English Nature, Roughmoor, Bishop’s Hull, Taunton, Somerset, TAI 5AA GENTIANELLA ULIGINOSA (WILLD.) BORNER (GENTIANACEAE) REDISCOVERED IN NORTH DEVON Dune Gentian Gentianella uliginosa (Willd.) B6rner is a European endemic and regarded as a “priority species” within the UK Biodiversity Action Programme. Rich (1996) reported previously unknown herbarium specimens collected at Braunton Burrows in North Devon (v.c. 4) prior to 1849 and in 1927 (all in BM), and suggested that it should be searched for again. Elsewhere in Britain it is known at five sites in south Wales (Kay 1972; Ellis 1983) and three on the island of Colonsay off western Scotland (Gulliver 1998; Rose 1998). On 28 August 1998 G. uliginosa was rediscovered at Braunton Burrows (SS/4.3) during a survey of all the slacks of this extensive dune-system. Over 130 plants of G. uliginosa were found in an area of several square metres in part of one dune-slack. A senes of voucher specimens was collected (NMW) and numerous photographs were taken. The plants identified as G. uliginosa were 2—4 cm tall, some reaching 6 cm, with one or two (rarely three) internodes; they all showed the long terminal pedicel (> one-half of total height to pedicel apex), and calyx lobes often of uneven size and spreading away from corolla, that are characteristic of this species. About 60 plants of Gentianella amarella (L.) Borner were growing intermixed with the G. uliginosa, and this species is widespread in numerous dune-slacks and grassland elsewhere at Braunton Burrows. Compared to G. uliginosa, G. amarella had much shorter terminal pedicels and mostly appressed calyx lobes that were all of similar size; they also included much larger plants (up to 21 cm tall) with more numerous internodes (4-10). However, at least seven plants of Gentianella associated with the population of G. uliginosa appeared intermediate between that species and the closely adjacent G. amarella in respect of pedicel length, number of internodes and calyx characters. The latter were judged to be hybrids between G. amarella and G. uliginosa, as were small numbers of plants seen during 24-30 August 1998 in other dune-slacks at Braunton Burrows accompanying G. amarella but not G. uliginosa. Hybrids with G. amarella are known from most colonies of G. uliginosa in south Wales, where they are reported to be fertile and to show “all grades of intermediacy” (Stace 1991), but they have not hitherto been reported from England. G. anglica (Pugsley) E. F. Warb. and its hybrids with G. amarella also occur in the same dune system (Rich et al. 1997), although in different slacks to G. uliginosa. Since Braunton Burrows is the only locality with both G. anglica and G. uliginosa, both of which grow alongside G. amarella, and the flowering seasons of all three species partly overlap, the possibility exists of other hybrids. Further investigations of the intermediate plants at Braunton Burrows are therefore planned. The G. uliginosa plants at Braunton Burrows were growing in almost closed cover of low vegetation (mainly c. 3 cm, tallest stems to 10 cm), on the nearly flat, humic sand of the floor of the dune-slack. The turf had numerous grasses and herbs, the commonest being Agrostis stolonifera L., Leontodon saxatilis Lam., Lotus corniculatus L., Holcus lanatus L. and Hydrocotyle vulgaris L.; only sparse Salix repens L. was present. Grazing by rabbits appears to be important at present in NOTES 429 maintaining the short sward at this location. However, rabbit numbers at Braunton Burrows have been much reduced over recent decades by myxomatosis and the consequent reduction in grazing has contributed to loss of much of the herb-rich turf for which this S.S.S.I. is famous (Breeds & Rogers 1998). Monitoring of Gentianella populations at Braunton Burrows is therefore needed to ensure timely management intervention where grazing pressure from rabbits declines. ACKNOWLEDGMENTS Field research at Braunton Burrows was carried out under contracts from English Nature and with access permission from the Ministry of Defence. Thanks are also due to John Breeds, Geraldine Holyoak and Martin Willing for assistance in the field, to Richard Gulliver for information on G. uliginosa, and to Tim Rich for checking voucher specimens. REFERENCES Breeps, J. & Rocers, D. (1998). Dune management without grazing: a cautionary tale. ENACT, managing land for wildlife 6: 19-22. E.uis, R. G. (1983). Flowering plants of Wales. National Museum of Wales, Cardiff. GULLIVER, R L. (1998). Population sizes of Gentianella uliginosa (Willd.) Boerner, Dune Gentian, on Colonsay (v.c. 102) in 1996. Watsonia 22: 111-113. Kay, Q. O. N. (1972). The dune gentian in the Gower Peninsula. Nature in Wales 13: 81-85. Ricu, T. C. G. (1996). Is Gentianella uliginosa (Willd.) Borner (Gentianaceae) present in England? Watsonia 21: 208-209. Ricu, T. C. G., Hotyoak, D. T., Marcetts, L. J. & Murpny, R. J. (1997). Hybridisation between Gentianella amarella (L.) Boerner and G. anglica (Pugsley) E. F. Warb. (Gentianaceae). Watsonia 21: 313-325. Rose, F. (1998). Gentianella uliginosa (Willd.) Boerner (Gentianaceae) found in Colonsay (v.c. 102) new to Scotland. Watsonia 22: 114-116. Stace, C. (1991). New Flora of the British Isles. Cambridge University Press, Cambridge. D. T. HoLyoak 8 Edward Street, Tuckingmill, Camborne, Cornwall, TR14 SPA RUBUS CAMPANIENSIS WINKEL EX BEEK (ROSACEAE) IN BRITAIN In 1977 I came across a large population of an unfamiliar glandular bramble with cupped pink flowers and distinctive leaf shape and prickle development on an overgrown old common near Emsworth, S. Hants., v.c. 11. A. Newton, to whom a specimen was shortly afterwards submitted, did not recognise it as any named species known to him, suggesting that it was perhaps a hybrid of R. sprengelii Weihe. Subsequent discovery of what was clearly the same bramble in two further localities in that district, however, rendered that suggestion less likely. Around 1982 a search of CGE brought to light a series of specimens matching this Hampshire plant from Hosey Common, near Westerham, W. Kent. v.c. 16, labelled as R. adornatiformis Sudre, a species recorded as frequent in that locality by Watson (1958). That name, however, was known to have been applied by Watson to British material doubtfully correctly; moreover, specimens from other localities in south-east England in BM, NMW and SLBI so determined by him are not only all different from his Hosey Common plant but mostly from one another as well. Edees & Newton (1988) were clearly well justified in relegating the taxon to an appendix listing names applied to British Isles Rubi dubiously at best. A further match was then made with two Dutch specimens in BM labelled as R. drymophilus Mueller & Lef., and a further one queried as R. granulatus Lef. & Mueller, all collected in 1951 ~ and sent by Kern & Reichgelt to F. Rilstone in an exchange. Again, however, neither of those names were judged to be applied correctly. The discovery that the plant was evidently present in the Low Countries nevertheless suggested that it would be worth sending specimens to the Belgian specialist in the group, H. Vannerom, and this was accordingly done in 1992. Vannerom at once recognised it as a bramble well known to batologists there. It had first passed for some years under 430 NOTES one or other of the two names used by Kern & Reichgelt and later put into wider currency by Beyerinck (1956), but, those having been found to be erroneous, the epithet campaniensis — after the Kempens district which bestrides the Dutch-Belgian border in the vicinity of Antwerp, where the bramble had turned out to be rather common — had been adopted instead. In the confident expectation that it would shortly be described as a new species by J. van Winkel, that name had already made at least one informal appearance in print (Vannerom 1986). Having examined a large number of sheets of the plant in 1986 in the Rijksherbarium at Leiden (L), I felt similarly safe in subsequently introducing the manuscript name into the British literature, attributing it to van Winkel (Allen 1996). In the event, however, van Winkel died before realising his intention, and it has fallen to van de Beek (1998) to publish the description. The holotype has been deposited in L, and an isotype donated to BM. Although van de Beek refers R. campaniensis to ser. Radula (Focke) Focke, the markedly variable expression of the armature, including its near-total suppression, seem to make it more appropriately placed in ser. Anisacanthi H. E. Weber (as in Allen 1996). In addition to the Kempens district the distribution is described as extending to Gelderland and Zeeland in the Netherlands, though much more thinly, and to Kent (locality unspecified) in England. However, as the following list of British exsiccatae indicates, its range in south-east England is actually much wider than_-that: v.c. 11, S. Hants.: abundant in chestnut plantations, Emsworth Common, SU/74.08, 23 July 1977 (BM), 19 June 1983 (herb. D.E.A., herb. H. Vannerom), 19 July 1992 (BM, BON), D. E. Allen. One patch, Havant Thicket, SU/715.113, 9 July 1977, D. E. Allen (BM). (The Emsworth Common population extends a short way into v.c. 13, W. Sussex). v.c. 15, E. Kent: east part of Denstead Wood, near Canterbury, TR/091.570, 14 July 1964, B. A. Miles, indet. (CGE). v.c. 16, W. Kent: Hosey Common, TQ/45.52, 13 Aug. 1905, anon. (but in handwriting of C. E. Britton), det. W. M. Rogers as R. pallidus var. leptopetalus forma, det. A. Newton 1977 as possibly R. praetextus (BM); 12 July 1934 (NMW), 21 July 1938 (CGE, SLBI), 6 July 1949 (CGE, NMW), 28 Aug. 1951 (SLBI), W. C. R. Watson, all as R. adornatiformis; 2 July and 10 Aug. 1954, 13 July 1955, 16 July 1957, W. H. Mills, all as R. adornatiformis (CGE); 2 July 1961, B. A. Miles as R. adornatiformis (CGE). v.c. 17, Surrey: Tilburstowhill Common, TQ/355.505, 16 July 1962, B. A. Miles (CGE). Hosey Common and Tilburstowhill Common are both on the Lower Greensand within 10 km of each other. It should be noted that specimens collected on the former by J. E. Woodhead in 1948 and 1951 and labelled R. adornatiformis (CGE), and on the latter by C. Avery in 1951 and labelled R. rotundifolius (SLBY, are not R. campaniensis but represent other, unnamed morphotypes. REFERENCES A.tLEN, D. E. (1996). Rubus L., in Brewis, A., Bowman, P. & Rose, F., eds The flora of Hampshire, pp. 147-160. Harley Books, Great Horkesley. BEYERINCK, W. (1956). Rubi neerlandici. Verhandelingen der Koninklijk Nederlandse Akademie van Wetenschappen, Amsterdam 51: 1-156. Epees, E. S. & Newton, A. (1988). Brambles of the British Isles. Ray Society, London. VAN DE BEEK, A. (1998). Nieuwe bramen uit de sectie Rubus uit het zuiden van het land. Gorteria 24: 19-30. VaNNEROM, H. (1986). Globale resultaten van een Rubus (sectie Rubus) — kartering in de provincie Antwerpen (Belgié). Dumortiera 34-35: 77-82. Watson, W. C. R. (1958). Handbook of the Rubi of Great Britain and Ireland. Cambridge University Press, Cambridge. D. E. ALLEN Lesney Cottage, Middle Road, Winchester, Hampshire, SO22 5EJ. NOTES 43] SEPARATION OF CAREX VULPINA L. AND C. OTRUBAE PODP. (CYPERACEAE) USING TRANSVERSE LEAF SECTIONS A rare plant in Britain, Carex vulpina is listed as vulnerable (Wigginton 1999) and is thought to have significantly declined in the last 20 years (Stewart ef al. 1994). In recognition of its threatened status an Action Plan has been produced (Anon 1995) as a framework to prevent further decline and plan for recovery. C. vulpina is superficially very similar to Carex otrubae and both are in the same subgenus Vignea (Stace 1997). The taxonomy of these two species in Britain was not elucidated until relatively recently (Nelmes 1939), but their separation remains problematical. Several morphological characters have been used to distinguish between the two species (Rich & Jermy 1998; Jermy, Chater & David 1982), but these can be variable, leaving the botanist with a degree of doubt over the plant in question. Hitherto, one of the most reliable characters has been the shape of the adaxial epidermal cells of the utricles, but even here there is some overlap between C. vulpina and C. otrubae and often material is difficult to place. Fundamental to the delivery of the conservation objectives is the ability to reliably distinguish this species from C. otrubae. Using material determined as C. vulpina, a comparison was made with C. otrubae, investigating a number of potentially useful morphological characters. It was noted that internal leaf anatomy, observed in transverse sections about half way up the leaf, differed markedly between the two species. Leaf sections were cut by hand with a double-edged razor blade using the tip of the forefinger as a cutting guide. Sections were mounted in water and observed under a compound microscope. An examination of many specimens, collected from Oxfordshire and Kent, showed that these differences remained constant. Leaf anatomy, as seen in transverse sections of the lamina, has been used extensively in taxonomic investigation of the Cyperaccac, including Carex, and many of the characters can be used for diagnostic purposes at the species level (Metcalfe 1971). Metcalfe’s book should be consulted for an explanation of the terms used in this note. The most useful anatomical feature in separating the two plants are the bulliform cells that overlie the midnb. In C. vulpina (Fig. 1) they are only slightly inflated, up to 70 um long (usually less), and are 3-tiered (sometimes 4-) and not strongly differentiated from the neighbouring chlorenchymatous cells. In C. otrubae (Fig. 2) the bulliform cells are strongly inflated, up to 85 um long, extending from the adaxial epidermis to the median vascular bundle, and arranged in a single tier forming a quite distinct group. In addition, the adaxial (upper surface) epidermal cells in C. vulpina are relatively small, about 20 um wide, whilst those in C. otrubae are larger, about 40 um wide. These and further differences are summarised below. FicurE | Carex vulpina, Otmoor, Oxfordshire 1998. FIGURE 2 Carex otrubae, Otmoor, Oxfordshire 1998. T.S. of keel region of leaf (sclerenchymatous tissue T.S. of kecl region of leaf (sclerenchymatous tissuc stippled). stippled). 432 NOTES Carex vulpina Carex otrubae bulliform cells numerous, not strongly inflated and at bulliform cells few, strongly inflated and single least 3-tiered tiered extending from adaxial epidermis to median vascular bundle adaxial epidermal cells small, 20 um wide adaxial epidermal cells about twice as large, 40 um wide air cavities within mesophyll + quadrate air cavities + elongate sclerenchyma associated with median vascular bundle — sclerenchyma associated with median vascular sits evenly in keel bundle offset sclerenchyma girders usually positioned abaxially sclerenchyma girders usually span width of lamina margin often incurved and filled with sclerenchymatous margin flat and sclerenchyma not in extreme leaf tissue in extreme lamina margin margin keel blunt keel sharp Leaf sections are easy to prepare and provide an unequivocal way of distinguishing between C. vulpina and C. otrubae. Furthermore, vegetative plants can be named obviating the need for inflorescences. Other large sedges sometimes grow with C. vulpina, including C. riparia Curtis and C. acutiformis Ehrh. These two species are morphologically distinct, particularly in ligule shape, but if there is any doubt they can be separated on leaf anatomy, both plants having a papullese abaxial epidermis seen most easily in transverse section. ACKNOWLEDGMENTS I would like to thank Roy Lambourne, Chris Pogson and Alan Showler for helping me find C. vulpina in the field; Paul Ashton for useful discussions, and particularly Colin Smith, (both Edge Hill University College) who kindly confirmed my observations by independently preparing and photographing leaf sections of both species. REFERENCES ANON (1995). Biodiversity: The UK steering group report. H.M.S.O., London. Jermy, A. C., CHaATER, A. O. & Davin, R. W. (1982). Sedges of the British Isles. B.S.B.I. Handbook No.1, 2nd ed. Botanical Society of the British Isles, London. MetTCALFE, C. R. (1971). Anatomy of the monocotyledons. Clarendon Press, Oxford. Nevmes, E. (1939). Notes on British Carices: IV. Journal of botany 77: 259-266. Ricu, T. C. G. & Jermy, A. C. (1998). Plant crib 1998. Botanical Society of the British Isles, London. Stace, C. A. (1997). New flora of the British Isles, 2nd ed. Cambridge University Press, Cambridge. STEWART, A., PEARMAN, D. A. & Preston, C. D., eds (1994). Scarce plants in Britain. Joint Nature Conservation Committec, Peterborough. Wicainton, M. J. (1999) Red Data Books of Britain: 1. Vascular plants. 3rd ed. Joint Nature Conservation Committec, Peterborough. R. D. PORLEY English Nature, Foxhold House, Crookham Common, Thatcham, RG19 SEL Watsonia 22: 433-441 (1999) A433 Book Reviews Dorset’s disappearing heathland flora. A. J. Byfield & D. A. Pearman. Pp. 37 (text) + 47 (annexes). Plantlife, London & Royal Society for the Protection of Birds, Sandy. 1996. £10.00, ISBN 9-780903-138987. This document does not, thank goodness, describe again the demise of Dorset’s heathlands. That particular story is already well-known, with large tracts of heathland lost — and those remaining much fragmented — due to agricultural “improvement”, forestry, mineral extraction and urban expansion. Rather, it reports on the status of rare vascular plants on surviving areas of heathland. The authors visited 390 heathland stands originally recorded by Professor Ronald Good as part of his remarkable floristic study of Dorset’s vegetation in the 1930s. They found 137 (35-1%) had been “destroyed”, while 253 (64-9%) had “survived” insofar as they still supported some kind of heathland or other semi-natural vegetation. In Good’s day each of these “surviving” stands Supported at least one rare species (defined by the authors as including national Red Data Book, nationally scarce and Dorset R.D.B. species, plus a few other “species of note’ — 41 species in all). The purpose of the new survey, carried out between 1990 and 1993, was to see how many of Good’s populations of these species were still extant. It was my Botany teacher at school who first pointed out to me that not finding a plant in a particular place was not the same as it not being there. Even so, the figures in this report are alarming, Of a total of 644 populations of rare species recorded by Good, the authors of the present study re-located only 163 (25-3%) — almost three-quarters had apparently been lost. This was more than double what one would have expected due to habitat loss alone. Take Radiola linoides as an example: Good recorded it in 74 stands, yet in the 38 of these stands still surviving in 1990-93 Byfield & Pearman could find Radiola in only two. For a few species the situation may not be quite as bleak as suggested in the body of the report. For example, Crassula tillaea is given as having suffered a 100% decline in Table 5, as it had disappeared from all seven of the sites in which Good found it in the 1930s; yet in Annex 3 there are records of it from four of Good’s stands in which he had not recorded it. Nevertheless, for most rare species the losses far outweigh the gains, for example: Anagallis minima, lost from eleven “surviving” stands and gained in just three; Cicendia filiformis, lost from nine, gained in two; Lycopodiella inundata, lost from 33, gained in none; and Radiola linoides, lost from 36, gained in four. As the report highlights, very few of these rare heathland species are found in heath plant- communities (sensu N.V.C.). Most occur in mire, or in ephemeral or early-successional vegetation within the heathland mosaic — along footpaths and cart-tracks, around the margins of seasonal pools and in puddles, in summer-parched sandy grasslands, on village greens and tightly grazed “lawns”. The authors consider that lack of management, or insufficient management — and especially the decline of grazing — has been the main reason for the observed declines, allowing patches of open species-rich vegetation to become increasingly overrun by Molinia, Juncus acutiflorus and ericoid shrubs. This report should be read by all those involved in the conservation of lowland heaths, not just in Dorset but elsewhere in the U.K. Its publication is timely, given the emphasis these days on biodiversity action plans and species recovery programmes. Management for rare species is a tricky business. Different species and groups of species — birds, reptiles, invertebrates and lower plants — have differing ecological needs. As this document makes crystal clear, on the Dorset heaths some major adjustments are required if the rare vascular plants are to survive. English Nature’s Wildlife Enhancement Scheme for the Dorset heathlands, established in 1994, has already succeeded in getting grazing back onto many areas (about 2000 ha and 25 sites by September 1998) — just in the nick of time, let us hope, and an initiative very much in line with this report’s view of what is needed. The authors do not beat about the bush: “in the long term we believe that only [through] the reinstatement of extensive pony and cattle grazing regimes over large tracts [of heathland] can the future of these plant species of nature conservation value be assured”. S. J. LEACH 434 BOOK REVIEWS Natives and Aliens — The wild flowers and trees of the Langdon Hills. Rodney L. Cole. Pp. 204. Basildon Natural History Society. 1996. £10.00, ISBN 0-9527849-0—4. Since its inception in 1968 the Basildon Natural History Society has achieved much in promoting a greater awareness of the wildlife that surrounds the new town of Basildon, and the publication of this work marked its thirtieth anniversary. An evocative introduction includes the author’s childhood memories of a time when it was safe for children to roam the local fields and woods unconsciously absorbing the natural world and, no doubt, coming home in a disreputable state. The first part of the publication covers not only the geography and geology of the Langdon Hills, but also an interesting account of its history and of how the landscape and the people have affected the area leading to the mixture of natives and aliens that make up the flora that is present today. Part II is a list of the plants present, with a very readable account for each species which shows the author has an eye for detail, an extensive knowledge of the area and that he understands the factors that have resulted in the plants present today. Colour plates of some of the rarer flowers are included, but there are only two depicting the landscape, both of historical interest; perhaps a modern one for comparison could also have been included. In summary, as a local Flora wnitten for the local people, it serves its purpose admirably. T. TARPEY Wildflower Safari: the life of Mary Richards. W. Condry. Pp. 237. Gower Press, Llandysul, Ceredigion. 1998. £17.99, ISBN 1—85902-558-7. There cannot be many B.S.B.I. members who have been the subject of a full-length biography, but when I finished reading this absorbing account, I felt that there would be few who would deny Mary Richards’ worthiness for the honour. Many readers will know of Richards only for her share with Peter Benoit 1n the “Contribution to the Flora of Merioneth” (2nd ed. 1963), but that 1s only a small part of her life. William Condry’s warm and sympathetic account, which 1s based on her diaries, supplemented by her papers, letters and the memories of many frends, begins in her childhood and bowls along well, past her marnage in 1907 and extensive terms abroad before and after the First World War. There is no mention of funds, but they must have been quite liberal to allow these trips plus extensive travel within the Bnush Isles as a keen member of the B. S. B. I. and the Wild Flower Society. Her husband died in 1941, when Mrs Richards was 56. Up to this point, and for the immediate post-war years, her life is energetic, seemingly impervious to the elements and thus like that of many others of our members. But in 1951, at age 66, and less than halfway through the book, she leaves for Central Africa, at first for an extended holiday, but soon to live there. Her life departs from the ordinary, and the prose of the author changes too. The account of early days is well-written, but because he is relying extensively on diaries it becomes just a little a series of unconnected events. Year follows year with selected highlights. With the arrival in Africa the pace rclaxes and the canvas broadens. I have done only a little botanising in Africa but the descriptions are right, the atmosphere is perfectly caught and with the accounts of the incredibly long days 1n the bush and then the ensuing hours of pressing and wniting up, the reader is filled with amazement at her energy, stamina and achievement. She continued long, arduous collecting trips for over 22 years, until just short of her 89th birthday, collecting around 20,000 specimens for Kew. We are not told how many new species were discovered by her and her African assistants, but she had one genus - Richardsiella (Poaceac) - and 28 species named after her. Each year she made trips to Bnitain, taking up her botany here again! William Condry says in his preface that years (20 in fact) have passed since the material was entrusted to him. We are fortunate that he completed the work just before his death and, indeed, saw the finished product. This is a nicely produced work with apposite photographs in colour and black and white. D. A. PEARMAN BOOK REVIEWS 435 Flora of County Dublin. D. Doogue, D. Nash, J. Parnell, S. Reynolds, P. Wyse Jackson (eds.). Pp. 560. Dublin Naturalists’ Field Club, Dublin. 1998. IR£25, ISBN 0—9530037—0-1. Special limited edition IR£ 150, ISBN 0-—9530037-1-X. It is said that the O’Connell monument in Dublin’s main street was designed by a committee and that 1t is none the worse for that. The new Flora of County Dublin has been written by no less than 19 members of the Dublin Naturalists’ Field Club, five of whom are credited with compiling and editing the work. The county has been well served by Nathaniel Colgan’s Flora of 1904 and a Supplement of 1961. The successful Flora of the Inner Dublin (P. Wyse Jackson & M. Sheehy Skeffington 1984) established a fieldwork team and a modus operandi which was used to good effect for the recent project. Fieldwork for the project continued until the early nineties though some more recent records are included. Dramatic urbanisation together with afforestation, peat erosion, drainage, roadmaking and coastal exploitation have all taken their toll on the “wild” county though much of interest and value remains. Sixtecn colour photos of good habitats are included in the Flora. Coastal habitats, gravel ridges, estate and semi-natural woodland, rivers and the mountains preserve the essential character of the flora as Colgan knew it. Praeger’s 1919 obituary of Colgan 1s reprinted, emphasising his outstanding contribution and his palpable presence to students of Dublin’s flora. A portrait of Colgan here would have been an enhancement. The introductory chapters on topography, climate, geology, soils, history of the study of the flora, botanical districts and 18 habitat accounts are written by specialists or particularly knowledgeable members of the Field Club. Some are short factual accounts, some didactic and some discursive and provocative. The account of the history of recording by Declan Doogue 1s a delight. There 1s a real sense of the author writing about kindred spirits. Some insecure references do not detract from this fine chapter: Wade’s discovery of Pastinaca at Finglas Bridge predates the founding of the Glasnevin Gardens; caution should attach to any assessment of David Orr’s record of Centaurium pulchellum, however “native” it looks now on the Bull Island. Orr perpetrated a large number of frauds, deliberately planting and subsequently finding rare plants or allowing others to find them! The Flora is intended to stand alone. Colgan’s 1904 records are repeated for the rarer plants. Tetrad dot maps are included for species with interesting or curious distributions, not wastefully for ubiquitous or very rare species. Colgan’s irritating but justifiable use of an appendix for casuals and aliens not fully naturalised is followed. The help of experts was enlisted for the critical genera, Rubus, Salix, Rosa, Potamogeton, Taraxacum and Chara. I wondered whether Sparganium erectum records might not be almost all subsp. microcarpum. “Circaea alpina, weed at Blackrock” was surely C. x intermedia. Epilobium obscurum, “occasional in base-poor wetlands” is also a common and troublesome garden weed, not at all particular about nutrient status. Mercurialis perennis may look native in several places in Ireland but is clearly spreading and a classic case of a relatively recent introduction. Colgan recognised it for what it is. Trifolium fragiferum survives at the Glasnevin pond by the River Tolka as McArdle reported it in 1902. Hydrocharis survived at Curragha at Icast until the mid-cighties. The Flora has been handsomely produced. What a joy to have it at last between covers. It 1s a credit to the Club. The “Committee” approach has delivered a fascinating selection of introductory essays and a meaty and detailed account of Dublin’s wild plants. Like Colgan’s Flora, which was its model and inspiration, and the aforementioned O’Connell monument, it will become another of Dublin’s worthy institutions. D. SYNNoTT 436 BOOK REVIEWS Dandelions of Great Britain and Ireland. B.S.B.1. Handbook No. 9. A. A. Dudman & A. J. Richards, ed. P. H. Oswald. Pp. 344. Botanical Society of the Bnitish Isles, London. 1997. £15.00, ISBN 0-901158-—25-9. This fine and scholarly Handbook is the fruit of a quarter of a century’s intensive study of the genus since Richards published his first attempt to bring order to British dandelions (The Taraxacum Flora of the British Isles 1972). The book is dedicated to the memory of the late Chris Howarth who played a major role in this revision. 235 species are described, a more than two-fold increase on the 132 in the earlier work since 34 of the latter are no longer recognised. There are excellent dichotomous keys, both to the Sections and to individual species or clusters of similar species. Bearing in mind the extreme difficulties of constructing keys to large apomictic genera this is as much as one can expect. There is also an ambitious multi-access key; it would be interesting to know just how useful such keys are in practice. 105 of the species are treated as “lead species”, with closely similar species being given the same number followed by a letter. The descriptions are full and detailed and points of comparison between related species are clearly indicated. The descriptions are augmented by silhouettes of whole plants and involucral drawings by the late Olga Stewart. There are interesting and helpful comments on apomixis, dandclion evolution and sources of variation, and invaluable advice on identification and on how, and how not, to collect. Distribution maps are provided for 178 species although these inevitably highlight the whereabouts of the relatively small number of dandelion enthusiasts. Sectional running heads are provided for the map section; it 1s a pity they are omitted in the main text. 96 (about 40%) of the species are thought to have been introduced and most of these are in the very large Section Ruderalia. The information on the extra-Bnitish distribution of the native species is patchy. There is a list (p. 15) of single species representative of phytogeographical areas. It would have been nice to have this expanded. The terms Western and Southern Atlantic hardly seem appropriate! One wonders whether, as in brambles, there are distinctive regional florulae. There is a cryptic reference (p. 7) to a “Taraxacum herbarium of the British Isles” built up by Richards and later augmented by Haworth but with no reference to its present whereabouts. It 1s in the care of A. A. Dudman. This 1s a first-rate Handbook, a worthy addition to the series and one of which the authors can be justifiably proud. G. HALLIDAY The Atlas Flora of Somerset. P. R. Green, I. P. Green & G. A..Crouch. Pp. xxiv + 292. Published by the authors. 1997. £25.00, ISBN 0—9531324—0-4. There have been two previous Somerset Floras, one by R. P. Murray in 1896 and the other by R. G. B. Roe in 1981. Neither was furnished with distribution maps which are such important features of the present work. An even more impressive feature, which becomes clear as the book 1s studied, is the remarkably complete coverage of a large county over the last ten years. The area includes 977 tetrads and the average number of species recorded per tetrad is 311, which exceeds that in almost all recent Floras. The authors are to be congratulated on the exceptional thoroughness of their fieldwork, as well as on their presentation of the data. Perhaps all authors of Floras would benefit from being twins! The initial information concerning methods, botanists, tetrad totals, geology and topography has been pruned to a minimum. It is supplemented by a summary of the 23 best botanising sites in the county, which at once demonstrates the breadth of the authors’ local knowledge. Throughout the text there are helpful hints as to where to find good colonies of the less common species. The text itself is set out on A4-sized paper in twin columns, with inset maps for about half of the 2300 or so taxa mentioned. This large number of taxa includes many aliens. There are 16 pages of colour plates in the central section of the book, all taken by the authors, and a comprehensive index. Although Somerset is conventionally divided into v.c.c. 5 and 6, its inland botany can be broadly described in three parts, West, Central and North. These parts are illustrated by many distribution maps in this Flora. To the west are the hills of Exmoor, with outliers in the Quantocks and the BOOK REVIEWS 437 Brendon Hills. These have high elevations, correspondingly high rainfall and acid soils. They are well provided with heath, bogs and Quercus petraea woodland, with plants such as Agrostis curtisii, Eriophorum vaginatum, Listera cordata, Nardus stricta, Oreopteris limbosperma, Sibthorpia europaea and Wahlenbergia hederacea. The authors conclude that Leucojum vernum, found in the lowlands here, is an ancient introduction. The central part includes the famous levels, where woods and even hedges are scarce, but the rhynes have a wealth of aquatics, e.g. Alisma lanceolatum, Ceratophyllum demersum, Hottonia palustris, Hydrocharis morsus-ranae, Utricularia vulgaris and Wolffia arrhiza. Here drainage and peatcutting have destroyed almost all the wet fen habitats, and Lathyrus palustris is all but extinct. If global warming does not trigger marine transgression during the coming century, this region may be further exploited to grow osiers for fuel. The northern part of Somerset is the least homogeneous, but it includes the botanically exciting limestone areas of Brean Down, Cheddar and the Mendips, with their relict populations of Carex montana, Dianthus. gratianopolitanus, Helianthemum apenninum, Koeleria vallesiana, Potentilla neumanniana, Saxifraga hypnoides, Thlaspi caerulescens and rare Hieracia. Of course some species have distributions which do not fit this oversimplified picture, examples being Cruciata laevipes and Rubia peregrina; the latter is frequent inland in this county. This Flora could be criticised for its omissions, the main being that it does not include that part of v.c. 6 which was part of the political county of Avon for the duration of the current survey. The reason was that a Flora for that ephemeral county was in preparation. The lack of sections on ecology and cryptogams can be remedied by future workers. The maps do not plot old records, but the text clearly states which species are thought to be increasing or decreasing. Conifers are perhaps less well covered than are other aliens, and the accounts of critical groups, such as Rubus, Hieracium and Taraxacum, are understandably incomplete. But these are minor quibbles. The Atlas Flora of Somerset can take its place as a first-rate modern County Flora, and is thoroughly recommended to residents and visitors alike. H. J. M. Bowen The Flora of Oxfordshire. J. Killick, R. Perry & S. J. Woodell. Pp. x11 + 386. Pisces Publications, Newbury. 1998. £45, ISBN I-874357-07-2. Oxfordshire has been a fortunate county, having had a series of Floras beginning in 1794 with Flora Oxoniensis by J. Sibthorp, updated by a series of later works. The last, published in 1927 was by G. C. Druce. Work started on this new one for v.c. 23 in 1968, the year Humphrey Bowen published his Flora of Berkshire for the adjacent v.c. 22 (which includes a substantial portion of present-day political Oxfordshire). However, the fieldwork behind this new volume has been executed extremely thoroughly over 28 years by an incredible number of botanists (duly acknowledged). Careful searching of herbarium and literature sources appears to have been carried out. There are some 90 pages of introductory material which give a good review of the topology, geology, palaeobotany (!), soils, climate, vegetation history and present-day communities (sumptuously illustrated in colour). A chapter entitled “The Vascular Plants of Oxfordshire’ actually gives details of some former Oxfordshire botanists, Floras, methodology of the recording exercise undertaken, inferences from the maps, dioccious species in the Oxfordshire flora and details of the structure and presentation of the species accounts. The lack of a decent topographical map 1s surprising. The main body of the Flora (212 pages) enumerates the species systematically with clear tetrad distribution maps for all but the most ubiquitous and those found in less than eight tetrads. Nomenclature follows Stace, including his usage of English vernacular names. Clear and helpful notes are given on abundance and ecology. Herbarium specimens for interesting, rare or unusual records are cited. Casuals are included and noted in smaller type, as also are species unrecorded but which might be expected as they occur just outside the border (e.g. Dactylorhiza traunsteineri (Sauter ex Reichb.) So6). This Flora unusually contains a substantial (60pp) account of the bryophytes by A. Roy Perry and the late Eustace W. Jones. This is in the same format with similar details and again with sumptuous colour illustrations (43 on 8 pages). No account of the lichens is provided. The work ends with a useful gazetteer, bibliography and index to scientific and vernacular names. 438 BOOK REVIEWS Modern county Floras are no longer produced by a single enthusiast and expert, but by a team. The three well-known and distinguished authors have not only been supported by the experts on bryophytes, soils, geology, etc., but by very many distinguished amateur and professional botanists to give an authoritative full treatment. The book, A4 size and heavy, is an altogether quality production. It is naturally going to be compared with the recent Flora of Cumbria by Geoffrey Halliday (University of Lancaster, 1997), which has set an incredibly high standard of both scientific content and production. The Flora of Oxfordshire comes up to this standard. All the illustrations — line drawings, colour paintings (four full-page by Andrew Brown) and colour photographs are extremely good and beautifully printed. The book is not onc to carry about, but for admiring and extensive reference 1n the library, office, study, etc. At £45 it is expensive, and it is to be hoped that botanists and plant-minded people will buy it. The authors, collaborators and publishers are to be congratulated on producing such a fine work. It is a shame we had to wait so long for it! Other workers currently producing county Floras must be intimidated about following these productions. S. L. Jury Scottish Wild Plants. Philip Lusby and Jenny Wnight with photography by Sidney J Clarke. Pp. 116. Royal Botanic Garden, Edinburgh. 1996. Hardback £19.95, ISBN 1—8722091-17-1. Paperback £12.95, ISBN 0-11—495802-5. This is a good book, ideal for dipping into for solace and inspiration in the winter and giving a fund of information for botanical excursions during the summer. It is the second in what is hoped will become a series on Scottish plant life and follows a broadly similar format to the book on Scottish orchids. In the foreword, David Ingram makes it clear that the aim of the book is to have professional botanists tell the story of some of Scotland’s plants in such a way as to foster concern on the part of an increasingly conservation-conscious public. The text consists of an introductory section which gives a brief history of the Scottish vegetation since the disappearance of the glaciers, an overview of the geographic elements represented in the Scottish flora and a discussion of the special features of the plants of the five broad regions of Scotland, often with reference to the plants included in the accounts. This is followed by a short section on the human influences on our flora, a sensible explanation of “rarity” and a too-bnef discussion of practical conservation measures. The “meat” of the book follows with an account of some 45 plants. There docs not appear to be a theme to the selection of species — except that it would seem that grasses and sedges are deemed not interesting or, more likely, not photogenic enough (or perhaps another book is already under consideration!). Many of the plants included are rare, some are quite frequent, but all have a good story well told. As we have come to expect, the photographs are wonderful, but the reproduction 1s far too small to do them justice although it will have kept production costs down. The frontispiece of Saxifraga oppositifolia is mouth-watering and the inside-cover of the hardback edition, with a wide-angle shot of Oxytropis halleri and Scilla verna, is stunning. The design of the cover 1s rather odd, with a fine picture of Coire Ba and the west end of Rannoch Moor superimposed by a picture of Moneses uniflora. The species accounts have a broadly consistent format detailing the history of the discovery of each plant, its British distribution, ecology, population biology and, where appropriate, conservation. As in the book on Scottish orchids, the distribution of each species is illustrated by a map of the botanical vice-counties in Scotland. For the rarer species, this scale 1s not very useful and can be very misleading — look at the distribution map of Phyllodoce caerulea and compare it with the distribution that is clearly explained in the text. Some explanation of the record of Moneses uniflora from Kintyre would also have been interesting! The amount of information condensed into a couple of pages is impressive as 1s its diversity, ranging from historical anecdotes to seriously complex sex-lives. This is the main strength of the book: the amalgamation, in a very readable format, of information for each species that would otherwise require many different sources. Inevitably some accounts generate as many questions as BOOK REVIEWS 439 they answer and there are a few factual errors. One wonders whether a few references within the text would really have been such a disruption. For most people who have looked at plants in Scotland there will be some personal connection with the text; I was delighted to read that the first record for Minuartia sedoides in Britain was from Ben Klibreck, where I saw it in abundance a few months ago. G. P. RoTHERO The Ferns of Britain and Ireland. C. N. Page. Second edition. Pp. 540. Cambridge University Press, 1997. Hardback £95, ISBN 0—521-—58380—2. Paperback £40, ISBN 0-521—-58658-5S. Anyone already interested in ferns has seen this book before; it was first published in 1982. It was then (and still is) the “bible” for all pteridologists — describing, in a loquacious but very readable style, (almost) all the pteridophytes found in the British Isles, and their habitats and associated plants. General chapters (including 19 maps of environmental parameters) and multi-access and chart keys give a useful background. However, for those who already have the first edition, I find myself asking: “Do the changes/additions warrant the outlay this book requires?” There are 100 or so extra pages; 20 on bracken alone with illustrations and descriptions of recent new taxa; three newly discovered Equisetum hybrids. Descriptions of taxa within Dryopteris affinis agg. have been brought into line with other recent authors but the hybrid D. x complexa, now included, is somewhat oversimplified. There is a useful discussion on Asplenium adiantum-nigrum and its serpentine form that is usefully recognised at subspecies level, [corrunense (Christ) Rivas-Mart., probably more correctly called subsp. silestacum Milde]. What have been added and indeed, enhance the book are some of Chris Page’s excellent black-and-white photographs of plants in the field (e.g. Equisetum sylvaticum on p. 469). The silhouettes (printed with greater contrast than in the first edition) of leaves and whole plants certainly help to give the “jizz” of a species but in some cases this has been over-done (e.g. for Selaginella selaginoides, 19 plants over two pages). This is one case where the paler, but greater detail, “xerox” reproduction in ed. 1 is more effective. Some things could have been improved, and indeed expected, if users’ feedback had been listened to. Alien ferns are not treated, even in passing, and this is particularly sad in the case of Azolla (we may have two species), and Equisetum ramosissimum, the presence of which in Somerset may be natural. Field botanists need to be acquainted with these plants. There is still a major gap in the absence of stem T.S. diagrams of the Equisetum hybrids, a group on which Page is undoubtedly a world authority. The small sketch-maps of the Bnitish Isles are infuriating and in many cases, among the rarer taxa where new records have been made, the fact is not recorded — or the map is changed and the text not, or vice versa. There are errors in authors of Latin names which hinders its use as a standard reference and it is clear that Page has revised some paragraphs but not others which would have benefited from it (e.g. taxonomic concepts on p. 5). Taxonomic concepts have changed a lot in the last 15 years. There is no doubt that Chris Page’s field observations are outstanding, and most of them are in this book. However, if he (and the publishers) are to achieve their aim of being really uscr-friendly, some serious editing to make the descriptive text comparative between close species or hybrids would be a great improvement. Perhaps this edition was revised in a hurry; and maybe the publishers should take some blame too. To answer the question I posed above: the keen fern enthusiasts will want this edition, regardless, but this reviewer feels that the full potential of both the subject and the author has not yet been fully realised. A. C. JERMY The Natural History of Pollination. M. Proctor, P. Yeo & A. Lack. Pp. 479. Harper Collins, 1996. Hardback £35, ISBN 0-00219905-—X. Paperback £16.99, ISBN 0—00219906-8. The study of pollination biology has come a long way since the publication of Michael Proctor & Peter Yeo’s The pollination of flowers (1973). So much so that the authors invited Andrew Lack to join them in writing an extensively revised and updated version of their earlier work. The result, 440 BOOK REVIEWS rather than going out as a second edition, has been published under a new title. The chief difference between the two works, apart from the enormous number of references to new studies, is that the new version adopts a rather more functional view of the phenomena involved and discussions are couched in terms of the costs and benefits to the various organisms involved. The subject matter covered, however, 1s similar between the two works, although the new version has a somewhat broader geographical scope, dealing with situations not only in the Bntish Isles, but also on a world-wide basis — as the topic dictates; this is particularly noticeable in the improved coverage of the various vertebrate pollination syndromes. Like its predecessor, this volume is packed with information and provides a fascinating and lucid discussion of the material in a style that 1s accessible to the intelligent layman as well as the professional. The topics that have benefited from updating are numerous and range from nectar composition and the chemistry of floral fragrances to late-acting self-incompatibility, the pollination biology of “primitive” angiosperms, and the ecology of pollination in plant communities; there is also a better account of the interrelationship between pollination biology and the genetical structure of populations. As a check on the extent to which new studies have been incorporated, I looked carcfully at the section on water-pollination and found that all those discoveries that I knew about, including internal geitonogamy and underwater outcrossing in the Callitrichaceae, bubble-pollination in Potamogeton and new observations of members of the Zosteraceae, were indeed dealt with. The treatment of orchid pollination is now more integrated, being addressed in a single chapter rather than being split between two as before, where Bnitish and European species were separated from the exotics; much more is also made of floral deception and brood-site pollination. Another innovation that I liked was the presentation of certain topics and definitions in “boxes” that can be read more-or-less independently of the text. Not surprisingly, in order to accommodate all the new material, some topics in the first edition have been omitted. Readers of this book who have an interest in the British & Insh flora will be disappointed to find that the short summaries of the pollination biology of selected plant species have gone, although some of the relevant data can still be found if searched for in the chapter on insect-pollinated flowers. Also gone is the discussion of speciation and reproductive isolation as promoted by different pollination syndromes. This is a pity because it is an important topic from an evolutionary point of view, although it is adequately covered in other books. The photographic illustrations, including eight composite coloured plates, of both flowers and pollinators are first-class, and complement the text beautifully. In short, this is an excellent book and I can recommend it unreservedly to anyone with an interest in the natural history of pollination. R. J. GORNALL Plant Crib 1998. T. C. G. Rich & A. C. Jermy, with the assistance of J. L. Garey. Pp. vii + 392. Botanical Society of the British Isles, in association with the National Museums & Galleries of Wales and the British Pteridological Socicty, London. 1998. Paperback £15.00, ISBN O0-901- 15828-3. The concept of Plant Crib 1998 can be traced back to Franklyn Perring’s "Blue Book’ entitled Hints on the determination of some critical species, microspecies, subspecies, varieties and hybrids in the British flora, which was published in Proc. B.S.B./. in September 1962 (Vol. 4, pp. 359-383) and also pre-published (March 1962) as a separate (price 3/6d) that was much used during recording for the Critical Supplement to the Atlas of the British Flora (1968). The more obvious predecessors of Plant Crib 1998 (Wigginton & Graham’s 1981 Guide (which was itself a revision of a 1976 work), Jermy & Camus’s 1987 The BM Fern Crib, and Rich & Rich’s 1988 Plant Crib) are outlined in the current work. This claims to have “been prepared to provide guidance with recording and identification of plants for the Atlas 2000 project’, but in fact it has turned out to be more than such a guidance, for it contains a lot of information which is not directly relevant to the Atlas 2000 work. For example, it carries sections on Taraxacum, Hieracium and Rubus, it includes many additional species (e.g. two in Gilia, three in Amelanchier) including a number not even known from the British Isles, and it covers several varieties (e.g. in Galeopsis and Fumaria), none of which will be included in Atlas 2000. BOOK REVIEWS 44] Plant Crib 1998 contains a large amount of very useful information that is certainly of enormous assistance in identification, and it should, if used wisely, greatly improve the accuracy of recording for Atlas 2000. The authors have performed an extremely valuable service in producing such a large and informative compendium in a very short timc. The limiting time factor explains the rather loose editing that is evident throughout. The authors have been forced largely to accept what they have received, so that there are many cases where the treatment could have been more usefully much shorter (e.g. Hymenophyllum, Oenothera) or much longer (e.g. Limonium, Melampyrum, Agrostis), where glaring gaps in coverage exist (e.g. Trichophorum subspecies), where important references appear to have been overlooked (e.g. Cerastium fontanum and Arenaria serpyllifolia), and where there are obvious inconsistencies (especially in nomenclature) and even errors (e.g. Deschampsia). Such imperfections are to be expected rather than criticized, but the user needs to be aware of them. A corrigenda sheet is now available. There are other important caveats for the reader to heed, all of which are hinted at by the authors in various places in the text. Firstly, this is not a book for beginners, but for the fairly experienced field botanist who will already have a good botanical vocabulary and will know how to distinguish each of the groups covered from taxa outside that group. Secondly, many of the “extra” characters used (i.e. those not usually considered diagnostic for the taxa concerned, but which are useful guides supplementing the strictly diagnostic features) are not absolutely diagnostic and must be used with great caution if they are not to mislead. The text must be scrutinized carefully to identify these, but sometimes they have found their way into the keys and tables, where they are liable to cause misidentifications. I personally found the accounts of Equisetum, Ranunculus subg. Batrachium, Alchemilla, Sorbus, Myriophyllum, Epilobium, Hedera, Callitriche and Carex particularly valuable, but other readers will have their own favourite sections. No field botanist will fail to find a great deal of value in the latest Crib. It is not so much a permanent reference work as a commentary on the current state of knowledge on British plant taxonomy, a situation which is still changing rapidly and which will call for rather frequent new editions (rather than supplements) in future years. As the authors acknowledge, much further research, testing and updating needs to be carried out before Plant Crib 199§ can be used as the basis for a Critical Flora of the British Isles. The latter is as distant a goal in 1998 as it was in the 1930s when first mooted. © A. SPACE ye see peri wat Bil neges es * } eee \ Saal, 5 = - 7 c ms i Tse df Ley ip) i iN = iT i = - ’ ; foi be oni J, i T rT i . qi i ' t y i * | pel” » » A i : - > oo x >} re é ~ ~ Watsonia 22: 443-449 (1999) 443 Obituaries JOHN HESLOP-HARRISON (1920-1998) One of the most innovative botanists of our time, John Heslop-Harrison will be sadly missed by his colleagues and admirers. He held three professorships (Belfast, Birmingham and Wisconsin), in addition to visiting professorships and many medals, fellowships and honorary degrees from learned societies and universities. In addition to being Director of the Royal Botanic Gardens, Kew (1971-76) he kept up a strong research programme, often with his wife Yolande and his son Pat, with some 300 publications in learned journals. I had the privilege of knowing him and working with him during his time at Birmingham (1960-67), and admired him greatly. In his carly days he worked on orchids, especially the dactylorchid populations in the Inner and Outer Hebrides, some of his work being published jointly with his father, J. W. Heslop-Harrison. He was also interested in the general aspects of Irish vegetation and post-glacial history. His book entitled “New Concepts and Flowering Plant Taxonomy” (1953) was an important landmark in plant evolution and classification concepts. In the late 1950s and early 1960s he was publishing, with his wife, Yolande on flowering-plant growth and organogenesis as shown in Cannabis sativa and other plants. He was also experimenting on apomixis and aspects of photoperiodic effect, and that of growth hormones, on sexuality and development. This led in the mid-1960s to investigations on pollen growth, cell walls and chloroplast structure, as well as sex expression in flowering plants. Thus, he was never content to describe what he saw but needed as a true scientist to investigate why it took place and the mechanisms which controlled such phenomena. These aspects of why and how things took place were published by him and with several of his research students and his wife in the mid- to late- 1960s. Thus, the what of taxonomy led to the why of physiology — what and why certain morphological and physiological phenomena took place. Many of these publications showed collaborative work with his students, particularly with Bruce Knox and Hugh Dickinson. He was always anxious to indicate his collaborators, though naturally the impetus came from Jack himself. In the 1970s Jack’s interest turned even more to the why and how of processes such as pollen-stigma interaction, incompatibility, and organelles, this work being the result of constant collaboration with Yolande. At this time he also wrote on the ever more important topic of species and ecosystem conservation, though at the same time, and into the 1980s, he was still investigating pollen growth and pollen-stigma interactions (again, with Yolande), as well as the phenomenon of heterostyly. Jack Heslop-Harrison’s contact with the B.S.B.I. will be well-known to older members, and particularly to those interested in dactylorchids and flora studies of Britain. His work and interests in so many other aspects of plant studies also will not be forgotten. Jack died of a heart attack on May 7th, 1998. Our deep sympathy goes out to his wife Yolande, his son Pat and other family members. SELECTED PUBLICATIONS Heslop-Harrison, J. (1983). Chromosomes, cladism and the new evolutionary debate. Kew chromosome conference 2: 313-322. Shivanna, K. R., Heslop-Harrison, J. & Heslop-Harrison, Y. (1983). Heterostyly in Primula 3. Pollen water economy: a factor in the intramorph-incompatibility response. Protoplasma 117: 175-184. Heslop-Harrison, J. (1982). The reproductive versatility of flowering plants: an over-view. In: Meudt, W. J. ed. Strategies of plant reproduction. BARC Symposium No. 6; Allanheld, Osmun, Totowa, 3018. Heslop-Harrison, J. (1982). Conservation of habitats and species — do we need to do anything? In: Hudson, W. J. ed. Biology, society and choice. Institute of Biology Charter Symposium, held in 1980, 71-78. Heslop-Harrison, Y. & Heslop-Harrison, J. (1981). The digestive glands of Pinguicula: structure and cytochemistry. Annals of botany 47: 293-319. Heslop-Harrison, Y., Heslop-Harrison, J. & Shivanna, K. R. (1981). Heterostyly in Primula 1. Fine-structural and cytochemical features of the stigma and style in Primula vulgaris Huds. Protoplasma 107: 171-187. 444 OBITUARIES Heslop-Harrison, J. (1979). Darwin and the movement of plants: A retrospect. Plant growth substances Madison, Wisconsin: 1-26. Heslop-Harrison, J. (1978). Summary and perspectives. Higher plants as monitors of environmental mutagens. National Institute of Environmental Health Sciences and the Energy Research and Development Administration, U.S.A. Eds. F. D. de Serres & M. D. Shelbey. Environmental health perspectives 27: 197-206. Heslop-Harrison, J. (1976). Introduction. In: Simmons, J. B. ef al., eds. Conservation of threatened plants. Plenum Press, 3-7. Heslop-Harrison, J. (1975). Man and the endangered plant. In: Pragnell, ed. International yearbook 1975: X11—XXVI. Heslop-Harrison, J. (1973). The plant kingdom: an inexhaustible resource? Transactions and proceedings of the Botanical Society of Edinburgh 42: |—15. Heslop-Harrison, J. (1971). Pollen: development and physiology. London: Butterworths. Heslop-Harrison, J. (1969). New concepts in flowering-plant taxonomy. Tokyo, Japan: Kaigai Hyoron-Sha. Heslop-Harrison, J. (1961). Arenaria norvegica Gunn, a species new to the Irish flora, in Co. Clare. Irish Naturalists’ journal 13: 267-268. Heslop-Harrison, J. (1961). Apomixis, environment and adaptation. Advances in Botany 1: 891-895. Heslop-Harrison, J. (1960). Temperature and vapour pressure deficit under drought conditions in some microhabitats of the Burren limestone, Co. Clare. Proceedings of the Royal Irish Academy 61:109-1 14. Heslop-Harrison, J. (1959). Apomictic potentialities in Dactylorchis. Proceedings of the Linnean Society of London session 170:174-178. Heslop-Harrison, J. (1959). Growth substances and flower morphogenesis. Journal of the Linnean Society botany 56: 269-281. Heslop-Harrison, J. (1959). Influence of day length on the breeding system of grasses. New Scientist 1959: 881-884. Heslop-Harrison, J. (1958). Darwin as a botanist. In: Barnett, S. A., ed. A century of Darwin. Heinemann, 267-297. Heslop-Harrison, J. (1957). Ring formation by Triglochin maritima in eastern Irish salt marsh. /rish Naturalists’ journal 12:1-6. Heslop-Harrison, J. (1957). The sexuality of flowers. New biology 23: 9-27. Heslop-Harrison, J. (1957). The physiology of reproduction in Dactylorchis. Auxin and the control of meiosis, ovule formation and ovary growth. Botaniska Notiser 110: 28—48. Heslop-Harrison, J. (1957). The hybridization of the common spotted orchid, Dactylorchis fuchsti (Druce) VermIn., with the marsh orchids, D. praetermissa (Druce) Vermln. and D. purpurella (T. and A. Steph.) VermIn. Proceedings of the Linnean Society of London 167:176-185. Heslop-Harrison, J. (1956). Some observations on Dactylorchis incarnata (L.) Vermil. in the British Isles. Proceedings of the Linnean Society of London 166: 51-82. Heslop-Harrison, J. (1956). Review: Botany of British Hills - Mountain flowers, by John Raven and Dr. Max Walters. Nature 178:1260. Heslop-Harrison, J. (1956). Dactylorchis traunsteineri Saut. in Co. Antrim. Irish Naturalists’ journal 12: 8-9. Heslop-Harrison, J. (1955). The conflict of categories. In: Lousley, J. E., ed. Species studies in the British flora. 160-172. Heslop-Harrison, J. (1955). Orchid hybrids in North Down. /rish Naturalists’ journal 11:1—4. Heslop-Harrison, J. (1955). Dactylorchis incarnata in the British Isles. Proceedings of the Linnean Society of London Session 166: 51-82. Heslop-Harrison, J & Webb, D. A. (1954). An atlas of plant-distribution in the British Isles. An appeal to Irish field botanists. The Irish Naturalists’ journal Xi: 1-3. Heslop-Harrison, J. (1954). A synopsis of the dactylorchids of the British Isles. Geobotanical Institute Riibel, Ziirich 28: 53-92. Heslop-Harrison, J. (1954). Review: The nature of Plant Species. Nature 174: 245. Heslop-Harrison, J. (1953). Genecology and orthodox taxonomy. Some theoretical aspects. Science progress 167: 484-494. Heslop-Harrison, J. (1953). Studies in Orchis 2. Orchis traunsteineri Saut. in the British Isles. Watsonia 2: 371-391. Heslop-Harrison, J. (1953). Some problems of variation in the British dactylorchids. South Eastern naturalist and antiquarian 58:14—25. Heslop-Harrison, J. (1953). New concepts in flowering-plant taxonomy. London, Heinemann. Heslop-Harrison, J. (1952). History of the British Flora. Advances in science 1952: 43-44. Heslop-Harrison, J. (1952). The North American and Lusitanian elements in the British flora. In: Lousley, J. E., ed. The changing flora of Britain. 105-123. Heslop-Harrison, J. (1952). The modern distribution of Irish plants in the light of postglacial history. Advancement of science 10: 42-44." OBITUARIES 445 Heslop-Harrison, J. (1952). A reconsideration of plant teratology. Phyton (Graz) 4:19-34. Heslop-Harrison, J. (1952). Phytogeographical account. Belfast in its Regional Setting. British Association handbook 6-8. Heslop-Hamson, J. (1952). Book Note: Wild Orchids of Britain, by V. S. Summerhayes. The journal of The Royal Horticultural Society 77: 143-144. Heslop-Harnson, J. (1951). Fresh aspects of Irish vegetational problems. II. /rish Naturalists’ journal 10: 145-149. Heslop-Harmison, J. (1951). The distribution of the Irish dactylorchids. Geobotanical Institute Riibel, Ziirich 25:100-113.. Heslop-Harmson, J. (1951). The history of Sphagnum. /rish Naturalists’ journal 10:152. Heslop-Harnson, J. (1951). Fresh aspects of Insh vegetational problems. I. Jrish Naturalists’ journal 10: 125-130. Heslop-Harmison, J. (1951). A comparison of some Swedish and British forms of Orchis maculata. Svenska Botaniska Tiddskrifi 45: 608-635. Heslop-Harmison, J. (1951). Review: Wild flowers of chalk and limestone, by J. E. Lousley. /rish Naturalists’ journal 10: 254. Heslop-Harrison, J. (1951). Review: The study of the distribution of Bnitish plants, being the report of the Conference held in 1950 by the Botanical Society of the British Isles, edited by J. E. Lousley. /rish Naturalists’ journal 10: 128. Heslop-Harmison, J. (1950). Notes on some Irish dactylorchids. Jrish Naturalists’ journal 10: 81-82. Heslop-Harmison, J. (1950). Orchis cruenta Mull. in the British Isles. Watsonia 1: 365-375. Heslop-Harrison, J. (1949). Orchis cruenta Mull., a new Irish marsh orchid. Jrish Naturalists’ journal 9: 329-330. Heslop-Harrison, J. (1949). Notes on the distribution of the Irish dactylorchids. In: Die Pflanzenwelt Irlands (The Flora and Vegetation of Ireland). Ergebnisse der 9. Intern. Pflanzengeographischen Exkursion durch Irland 1949. Redigiert von Werner Ludi — Verlag Hans Huber Bern. Zurich: Geobotanischen Institutes Rubel, 100-113. Heslop-Harmison, J. (1949). Intersexuality in Irish Willows. /rish Naturalists’ journal 10: 269-272. Heslop-Harmison, J. (1949). The dactylorchids of N. W. Donegal. /rish Naturalists’ journal 10: 291-297. Heslop-Harrison, J. (1949). Review: Studies in late quaternary deposits and flora-history of Ireland, by Knud Jessen. Proc. Roy. trish Acad. 52: 85-290. Heslop-Harrison, J. (1949). Maps of the biological subdivisions of Ireland. The Irish Naturalists’ journal 9: 331-333: Heslop-Harrison J. W. & Heslop-Harrison, J. (1949). Notes on the flora of the Isles of Lewis, Harris, Killegray and Ensay. Transactions and proceedings of the Botanical Society of Edinburgh 35: 132-156. Heslop-Harmison, J. (1948. Field studies on Orchis. The structure of Dactylorchid populations on certain islands in the Inner and Outer Hebrides. Transactions and proceedings of the Botanical Society of Edinburgh 35: 26-66. Heslop-Harnison, J. & Clark, W. A. (1948). Noteworthy plants from Great and Little Bernera (Lewis), Pabbay and Berneray (Harris), and the Uig district of Lewis. Proceedings of the University of Durham Philosophical Society 10: 214-221. Heslop-Harrison, J. & Heslop-Harrison, J. W. (1948). The vascular plants of Stuley Island, the Isles of Grimsay and Raasay, with some remarks on the flora of Benbecula, South Uist and Barra. Proceedings of the University of Durham Philosophical Society 10: 499-515. Heslop-Harrison, J. (1946). The flora and fauna of the Western Isles of Scotland and their biogeographical significance. Proceedings of the Belfast Natural History and Philosophical Society 1946: 87-96. J. HAWKES OLGA MARGARET STEWART (1920-1998) Olga Margaret Stewart (née Mounsey), who died on 6 August 1998, was inextricably linked with the botany of Scotland. She had been a member of both the B.S.B.I. and the Wild Flower Society since 1965, and soon became close friends with Mary McCallum Webster, who had just settled in Morayshire. Olga was quick to learn from her mentor and, as her knowledge increased, so did her enthusiasm. In 1975 she was invited to become Recorder for the Stewartry of Kirkcudbright (v.c. 73), and it is for her thorough work there that most of her British botanical friends will remember her. 446 OBITUARIES She was bom in Edinburgh on 1 July 1920. Her father was an Edinburgh lawyer, James L. Mounsey, and her mother a Canadian from Nova Scotia. Her first school was in Edinburgh and at the age of 12 she moved to a boarding school in Kent, where she had a successful career, winning several academic prizes and representing the school at hockey, lacrosse and tennis, and ending up as joint Head Girl. She returned to Edinburgh to study architecture at the Art College, and at the end of her first year was on holiday in Nova Scotia visiting her grandfather when World War II broke out. It was decided that she should stay there, and she enrolled into Dalhousie University in Halifax, where she found herself the only female studying Engineering. Her professors soon noticed her drawing capabilities and, after a year, they offered her a job with the National Research Council of Canada in the Naval Dockyard at Halifax. When the first German acoustic mine was discovered and defused off the coast of Nova Scotia, it was Olga who was given the task of producing sectional drawings of it for military scientists to study. In 1943, in spite of hostilities both in the North Atlantic and Western Europe, Olga decided to return to the U.K. and was given a research job with the Royal Navy in Edinburgh. Three years later, on 28 November 1946 in Edinburgh, she married Frank Stewart, a pre-war fnend who had spent the last five years as a Prisoner of War in Germany; they enjoyed a very happy life together for over 51 years. Frank was yet another Edinburgh lawyer, so it was there, in between bringing up four children, that Olga started on her botanical ‘career’. This had a curious beginning. In 1947 she and Frank went on their first holiday together, to Aviemore, where, as she wrote later, (1983. Wild Flower Society Magazine 39: 15) “While climbing a hill, rather slowly and rather pregnant I collected fiowers on the way and sat down to draw them ... So began my passion to draw practically every plant, wild and naturalised that I have seen since”. Her reputation as an artist soon spread and she was asked to draw black-and-white flower illustrations for books and journals. Examples are to be found in McCallum Webster’s Flora of Moray, Nairn & East Inverness (Aberdeen University Press, 1978), and the late Princess Grace of Monaco’s book, My Book of Flowers (Doubleday, 1980), and many other articles on British plants. Most recently (1997), she drew the involucres of Bntish dandelions in the B.S.B.I. Handbook 9, by Andrew Dudman and John Richards; to use the authors’ own words, “She has graced the present volume with delightful and accurate drawings ... and we thank her for the dedication and hard work that these have entailed”. She also left, unpublished, a number of hawkweed drawings, and a collection of water-colour drawings of more than 3000 Bnush plants, which is indeed unique, and well worth publishing in its own night. Her drawing of the bluebell has become well-known being incorporated into the Society’s logo. It is Olga’s botanical fieldwork in Kirkcudbrightshire for which she will be best remembered. She and Frank had enjoyed summer holidays around New Abbey for some years and in 1962 had the opportunity to build a holiday home, ‘West Maryfield’, on the north-east of the village. She had got to know the local flora well, so on Dr Humphrey Milne-Redhead’s demise in 1974, she was the obvious choice to carry on as B.S.B.I. Recorder for the vice-county. Characteristically, she took up her duties seriously and enthusiastically, scouring the area from the top of the Galloway Hills to the Solway, often three days a week between March and October, sometimes alone and sometimes with friends. Her Check List of the plants in Kirkcudbrightshire was published in 1990, by the Dumfries & Galloway Natural History and Antiquarian Society, of which she was an active member. Her card index with detailed records and her maps at tetrad level destined for later publication will be completed, we must hope, and brought to fruition. Stimulated by the lack of expertise available to identify problems in the smallreed (Calamagrostis spp.) complex she took up the cudgels and studied plants from elsewhere in Britain and Europe (including type specimens) in the National Herbaria at Edinburgh and the Natural History Museum, London. She became the B.S.B.I. Referee for the group and the characters she listed as useful for the five species in 1988 were re-published in the Plant Crib 1998, as being still the authontative account. Paintings, publications and records are lasting tributes to some of her endeavours which can be enjoyed by those that study plants in the future. But to the large coterie of friends she leaves behind ~— the many who enjoyed hospitality at West Maryfield, on their way from the South to northern Scotland — it will be those warm and happy memories of botanical forays, long walks to re-find a rare plant, or to check the status or identity of an old record; memories of botanical discussions in the kitchen, poring over maps and checking grid references over pre-dinner gins and tonic, unobtrusively administered by Frank, whose silent support for all she did allowed Olga to develop OBITUARIES 447 her hobby to the full. After dinner, there would be the checking of taxonomic queries with lens or microscope, and Olga would carefully put those little vouchers, invariably shrivelled, into the plant-press which, when dry, would be sent to specialists at home or abroad to confirm the record; or taken to the R.B.G. at Edinburgh to compare with herbanum material. Then, later in the evening, replete in body and relaxed in mind, her guests would retire, leaving Olga sketching a wanted plant before the petals fell or dried. Next morning she was the first up, and breakfast (and packed lunch!) would be ready by the time we came down. If Olga thought a species should be in ‘her area’ she would not rest until it was found, and confirmed. Invariably she succeeded, or manoeuvred her botanical friends into the night habitat so that they could “find” it. I don’t know how many scraps of Dryopteris dilatata I was shown before the true mountain buckler fern (D. expansa) was found, but found it was! The challenge at the time of her death was to find the hybrid between Isoetes echinospora (not common in Kirkcudbnghtshire) and /. lacustris. This enthusiasm for something new was infectious, and her presence at any meeting was welcomed with real pleasure. The B.S.B.I. Exhibition Meetings in Scotland were never, or rarely, missed, and for many years she also regularly attended — and exhibited at — the Society’s Annual Meeting in London. On the rare occasion when she did not, her absence was the concerned comment of many who had looked forward to seeing her. Over a period of 50 years she acquired a great knowledge of flowers, and she was, without doubt, one of the best field botanists in Scotland, if not Britain. She was concerned about nature conservation and encouraged landowners and other environmental managers to do their best for plants. As a Regional representative on the B.S.B.I. Conservation Committee for S.W. Scotland, she was particularly concerned about the water management at Loch Ken, and produced a thorough survey of the area (1988. Transactions of the Dumfries & Galloway Natural History & Antiquarian Society 63: 1—4) to draw attention to the plight of the pillwort (Pilularia globulifera). Olga was always interested, too, in other people, young and old alike, and from time to time worked for charities helping elderly people. She helped run a Badminton Club for young people in Edinburgh in the 1960s but her main sporting love was curling. She joined the Edinburgh Ladies Curling Club in the mid-1950s and, as 1n everything she took up, she soon became an expert. Olga won many competitions and, in 1967, was chosen to go on a curling tour of Western Canada, from Winnipeg to Victoria, as one of 20 ladies representing Scotland. She was forced to retire from curling only two years ago, after slipping and hitting her head on the ice. She had the ability to impart her great knowledge of botany to beginners in her own unassuming way; She always saw the good side of a person’s character and never criticised others. Her integrity and generosity of spirit is something that all of us who have had the good fortune to know her will remember most. Olga leaves husband Frank, daughter Rosemary, and sons Alan, John and Nick, the last of whom, having graduated in Geology at Cambridge, followed in her footsteps and is now a professional botanist in his own night. They, and their families, have lost a loving wife, mother and grandmother and our sympathies are with them. There are also botanical fnends who have enjoyed many a happy day exploring Kirkcudbright with Olga. We too have lost a most knowledgeable and lovable field companion. A. C. JERMY KARL-HEINZ RECHINGER PHIL. DOC. (VINDOB.) FMLS (1906-1998) Karl-Heinz Rechinger, author of Flora Aegaea, founder, editor and for many parts also author of Flora Iranica, died in Vienna on 30 December 1998, aged 92. He has increased our knowledge of the flora and vegetation of the Balkan peninsula and South-west Asia more than anyone in this century. Since 1938 he has been a member of the B.S.B.I., and was later elected to honorary membership. The combination of three qualifications made Rechinger a most remarkable man. He acted for almost 35 years as director of the Department of Botany at the Natural History Museum in Vienna and for eight years served as “Erster Direktor” of that institution. At the same time he was also a prolific scientist who left behind a vast oeuvre of the highest merit with scores of new taxa described, of which many have stood the test of time. Thirdly, Rechinger was an extremely active plant collector who repeatedly travelled to some of the remotest corners of the world. 448 OBITUARIES When the first instalment of Watsonia was published in January 1949 it contained a contribution by Rechinger which had first been presented at a meeting of the British Association in Dundee. Entitled “Lines of evolution and geographical distribution in Rumex subgen. Lapathum’’, its introduction offers a good insight into his mind: “The docks, having no showy flowers, do not awake aesthetic feelings as do, for instance, brightly coloured flowering plants or plants distinguished by a particular habit. The more intimate beauty which lies in the various shades of red, brown and yellow of the ripe fruiting panicles, contrasting with the dull green willows accompanying the river banks is not so obvious. Nevertheless it belongs undoubtedly to the general impression of a landscape in autumn. On the other hand, when looked at with a certain attention the amazing variability in size and shape of the inner perianth segments at the fruiting period offers a great deal of pleasure’. Pleasure meant for Rechinger travelling and collecting plants in dry habitats, with the Aegean area and the Near East being his favourite hunting grounds. Travelling in Iran, Iraq, Afghanistan and Pakistan was as that time an uncomfortable experience and required much dedication. He has brought home more than 100,000 collections, mainly flowering plants, many in several duplicates, and almost always deposited the first set in the Natural History Museum in Vienna. Pleasure meant for Rechinger analysing the endless variety of plant life, and descnbing it with great care and accuracy, Cousinia, Rumex and Salix being among his special friends. But pleasure also meant for Rechinger synthesising his observations and those of others in Floras, of which he produced three: Flora Aegaea, his first magnum opus, published in 1944; Flora of Lowland Iraq, printed in 1964; and his second magnum opus, Flora Iranica, of which the first fascicle came out in 1963. So far 173 fascicles have been published, a most remarkable achievement, with the treatment of a single family, a single tribe, and a single, albeit large, genus (Astragalus) still outstanding. A eulogy in Annalen des naturhistorischen Museums in Wien 75:1-16 (1971) and a paper in Proceedings of the Royal Society of Edinburgh 89B: 3-5 (1986) are major sources of information about his numerous publications. | Rechinger visited England and Scotland several times, maintaining long-term contacts with several botanists, notably at Edinburgh and Kew. Through his close co-operation with Peter Davis, editor of Flora of Turkey and the East Aegean islands, he strove for consistency with Flora Iranica despite there being, at times, varying perceptions of taxonomic rank. For many members of the B.S.B.I. he was their first contact when dealing with queries concerning the very rich holdings of the Natural History Museum in Vienna. There was a considerable involvement of British botanists in Flora Iranica: the account of Papaveraceae was written by J. Cullen, Rutaceae by C. C. Townsend, Guttiferae by N. K. B. Robson, Capparidaceae, Aizoaceae and Molluginaceae by I. C. Hedge and J. Lamond, Gramineae by N. L. Bot, several mainly aquatic monocot families by J. E. Dandy, and Balsaminaceae by C. Grey-Wilson. Several British botanists (A. J. C. Grierson, I. C. Hedge, J. Lamond, J. A. Ratter and P. F. Yeo) also contributed generic treatments to individual families, and the most recently published two fascicles of Flora Iranica were edited by I. C. Hedge instead of the ageing Rechinger. Some of his expeditions to Afghanistan, Pakistan and Iran were done in the company of two Scottish botanists, I. C. Hedge and J. Lamond. Rechinger was one of the original advisory editors to the monumental Flora Europaea project, and although his original contributions were restricted to accounts (notably Salix and Rumex) in the first volume, published in 1964, his works on the Flora of Greece continued to provide a major foundation for studies of the flora of that country. Rechinger grew old enough to be able to reap what he had sown, and he could reflect many of his undertakings in two autobiographical papers transmitting something of the flavour and fascination of travelling in the Levant and the Near East. They were published in the Annales Musei Goulandris in 1978 and the Davis & Hedge Festschrift in 1989 by Edinburgh University Press. Music was also very important to him, both playing the piano and listening to classical music, a pleasure from which deafness deprived him late in life. Apart from many other distinctions like the Omajou Order Third Class from the Shah of Iran, Rechinger was honoured by being elected Foreign Member of the Linnean Society of London in 1966 (restricted to 50 botanists and zoologists), and was appointed an Honorary Fellow of the Royal Society of Edinburgh in 1987. His many friends will remember his benevolent, gentlemanly nature and his puckish yet kindly sense of humour. He is survived by his widow, Wilhelmina, without whose support in all facets of his life he would never have achieved so much. OBITUARIES 449 without whose support in all facets of his life he would never have achieved so much. His was a very rich and remarkable life, reaching the utmost end of what nature has accorded to man. British botany, and Bntish botanists with their colleagues on the Continent and worldwide, will miss Karl-Heinz Rechinger. H. W. Lack — SAO Ne aa = 7 West Comwall Scilly East Comwall South Devon North Devon South Somerset North Somerset North Wiltshire South Wiltshire Dorset Isle of Wight South Hampshire North Hampshire . West Sussex East Sussex . East Kent West Kent . Surrey . South Essex North Essex Hertfordshire . Monmouthshire . Glamorgan . Breconshire . Radnorshire . Carmarthenshire Dumfnesshire . Kircudbnghtshire Wigtownshire Ayrshire Renfrewshire . Lanarkshire Peeblesshire Selkirkshire Roxburghshire . Berwickshire . East Lothian . Midlothian West Lothian Fife . South Kerry . North Kerry . West Cork . Mid Cork . East Cork . Co. Waterford . South Tipperary . Co. Limerick . Co. Clare H10. Hil. H12. H13. H14. North Tipperary Co. Kilkenny Co. Wexford Co. Carlow Laois NAMES OF VICE-COUNTIES IN WATSONIA DiI. 22s 723), 24. . East Suffolk . West Suffolk . East Norfolk . West Norfolk . Cambnidgeshire ENGLAND Middlesex Berkshire Oxfordshire Buckinghamshire Bedfordshire . Huntingdonshire . Northamptonshire . East Gloucestershire . West Gloucestershire Herefordshire . Worcestershire . Warwickshire . Staffordshire . Shropshire . Shropshire . South Lincolnshire WALES . Pembrokeshire . Cardiganshire . Montgomeryshire . Merionethshire Caermarvonshire SCOTLAND Stirlingshire West Perthshire . Mid Perthshire East Perthshire Angus . Kincardineshire . South Aberdeenshire . North Aberdeenshire Banffshire . Moray Easterness . Nairnshire . Westerness . Main Argyll IRELAND . South-east Galway . West Galway . North-east Galway . Offaly . Co. Kildare . Co. Wicklow . Co. Dublin . Meath . Westmeath . Co. Longford . Co. Roscommon . East Mayo . West Mayo . Co. Sligo 54. 2b) 55b. 56. Sie 58. oy): 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 69b. Uo, Ale 50. Silt 52) 99. 100. 101. 102. 103. 104. 105. 106. 107. 108. 109. 110. 111. J, H29. H30. H31. H32. H33. H34. H35. H36. H37. H38. H39. H40. North Lincolnshire Leicestershire Rutland Nottinghamshire Derbyshire Cheshire South Lancashire West Lancashire South-east Yorkshire North-east Yorkshire South-west Yorkshire Mid-west Yorkshire North-west Yorkshire Co. Durham South Northumberland North Northumberland Westmorland Furness Cumberland Isle of Man Denbyshire Flintshire Anglesey Dunbarton Clyde Isles Kintyre South Ebudes Mid Ebudes North Ebudes West Ross East Ross East Sutherland West Sutherland Caithness Outer Hebrides Orkney Shetland Co. Leitnm Co. Cavan Co. Louth Co. Monaghan Fermanagh East Donegal West Donegal Tyrone Co. Armagh Co. Down Co. Antnm Co. Londonderry BOTANICAL SOCIETY OF THE BRITISH ISLES 5 phe ea The B.S.B.I. traces its origin to the Botanical Society of London founded in 1836 and has a membership of 2,850. It is the major source of information on the status and distribution of British and Irish flowering plants and ferns. This information, which 1s gathered through a network of county recorders, is the basis for plant atlases and of publications on rare and scarce species and is vital to botanical conservation. The Society published Atlas of the British flora in 1962 and revised editions and is working towards a new atlas, to be completed in 2000. It organises plant distribution surveys, publishes handbooks on difficult groups of plants and has a panel of referees available to members to name problematic specimens. The B.S.B.I. arranges conferences and field meetings throughout the British Isles and, occasionally, abroad. It welcomes as members al! botanists, professional and amateur alike. Details of membership and any other information about the Society may be obtained from: The Hon. General Secretary, Botanical Society of the British Isles, c/o Department of Botany, The Natural History Museum, Cromwell Road, London, 5W7 5BD. BSBI Handbooks Each Handbook deals in depth with one or more difficult groups of British and Irish plants. No.1 Sedges of the British Isles A.C. Jermy, A.O. Chater & R.W. David. Revised edition; 1982. 272 pp., with descriptions, line drawings and distribution maps for all 73 species of Carex. Paperback. ISBN 0 901158 05 4. No.2 Umbellifers of the British Isles T.G. Tutin. 1980. 200 pp., with descriptions and line drawings of 73 species of Apiaceae (Umbelliferae). Paperback. ISBN 0 901158 02 X. No.3 Docks and knotweeds of the British Isles J.E. Lousley & D.H. Kent. 1981. 208 pp., with descriptions and line drawings of about 80 native and alien taxa of Polygonaceae. Paperback. Out of print. New edition with distribution maps in preparation; orders recorded. No.4 Willows and poplars of Great Britain and Ireland R.D. Meikle. 1984. 200 pp., with descriptions and line drawings of 65 species, subspecies, varieties and hybrids of Salix and Populus. Paperback. ISBN 0 901158 07 0. No.5 Charophytes of Great Britain and Ireland J.A. Moore. 1986. 144 pp., with descriptions and line drawings of 39 species and varieties of Characeae and 17 distribution maps. Paperback. ISBN 0 901158 16 X. No.6 Crucifers of Great Britain and Ireland T.C.G. Rich. 1991. 344 pp., with descriptions of 148 taxa of Brassicaceae (Cruciferae), 129 of them with line drawings, and 60 distribution maps. Paperback. ISBN 0 901158 20 8. No.7 Roses of Great Britain and Ireland G.G. Graham & A.L. Primavesi. 1993. 208 pp., with ibscupuGns and line drawings of 13 native and nine introduced taxa of Rosa, descriptions of 76 hybrids, and 33 maps. Paperback. ISBN 0 901158 22 4. No.8 Pondweeds of Great Britain and Ireland C.D. Preston. 1995. 352 pp., with descriptions and line drawings of all 50 species and hybrids of Potamogeton, Groenlandia and Ruppia, most of them with distribution maps; detailed introductory material and bibliography. Paperback. ISBN 0 901158 24 0. No.9 Dandelions of Great Britain and Ireland A.A. Dudman & A.J. Richards. 1997. 344 pp., with descriptions of 235 species of Taraxacum, most of them illustrated by silhouettes of herbarium specimens; drawings of bud involucres of 139 species and 178 distribution maps. Paperback. ISBN 0 901158 25 9. Other publications English names of wildflowers J.G. Dony, S.L. Jury & F.H. Perring. 1986 (2nd ed.). 126 pp. Recommended English names for British and Irish vascular plants (Latin-English and English-Latin). Paperback. ISBN 0 901158 IS 1, List of vascular plants of the British Isles D.H. Kent. 1992. 400 pp. The same nomenclature and sequence as in Clive Stace’s New Flora of the British Isles (1991, 1997), with selected synonyms. Paperback. ISBN 0 901158 21 6. Supplied with five errata lists and Supplement I (December 1996) of 36 pp. Alien plants of the British Isles E.J. Clement & M.C. Foster. 1994. 616 pp. Lists 3,586 recorded non-native species (of which 885 are established), with English names, frequency of occurrence, status, areas of origin, location of voucher specimens, references to published descriptions and illustrations, and selected synonyms. Paperback. ISBN 0 901158 23 2. Alien grasses of the British Isles T.B. Ryves, E.J. Clement & M.C. Foster. 1996. 234 pp. A companion volume to the last, listing over 700 non-native grasses. Includes keys to bamboos and eight of the larger and more difficult genera and 29 pp. of illustrations. Paperback. ISBN 0 901158 27 5. Plant crib 1998 T.C.G. Rich & A.C. Jermy. 1998. 400 pp. An identification guide for some 325 ‘difficult’ taxonomic groups, with explanations, keys and illustrations of plant details. A4 paperback. ISBN 0 901158 28 3. British Red Data Books I Vascular plants M.J. Wigginton, ed. 1999 (3rd ed.). 468 pp. Up-to-date information and maps for 408 taxa (including 118 microspecies) regarded as threatened in Great Britain, with details of their habitats and associated species, maps showing their pre-1970, 1970-1987 and post-1987 records by 10-km squares, and 1-km square frequency maps for 63 taxa. Hardback, published by JNCC, Peterborough. ISBN 1 86107 451 4. Scarce plants in Britain A. Stewart, D.A. Pearman & C.D. Preston, comp. & ed. 1994. 518 pp. Accounts of 254 nationally scarce taxa (occurring in 16-100 10-km squares in Great Britain) and of 71 taxa formerly thought to be so, with details of their habitats and associated species, reproductive biology, changing British distribution and world range; updated distribution maps. Hardback, published by JNCC, Peterborough. ISBN 1 873701 667. Aquatic plants in Britain and Ireland C.D. Preston & J.M. Croft. 1997. 365 pp. Summarises the distribution, habitat and reproductive biology of 200 aquatic plants in 72 genera, with 200 distribution maps and 72 line drawings. Hardback, published by Harley Books, Colchester. ISBN 0 946589 55 0. Available from the official agents for BSBI publications, F. & M. Perring, Green Acre, Wood Lane, Oundle, Peterborough PE8 4JQ (Tel: 01832 273388 Answerphone: 01832 274892. Fax: 01832 274568. e-mail: perring @btinternet.com). : ie ak a 7 eo . >t - " rh : .. piety '9 wh PRED 2: sv het’ Ai a4 a Ph i & Soe, on ora vl! ne al? Oe ye o. es py ih f ie | buts et! oe t - ' fe 1 3 ‘ j ‘en et : i : = % , q ~~, > 5 = mere i re } ‘ i = on he { iy ti tied head 1 bd - - & - — = ~ - my 5 INSTRUCTIONS TO CONTRIBUTORS Scope. Authors are invited to submit Papers and Notes concerning British and Irish vascular plants, their taxonomy, biosystematics, ecology, distribution and conservation, as well as topics of a more general or historical nature. Authors should consult the Hon. Receiving Editor for advice on suitability or any other matter relating to submission of manuscripts. Papers and Notes must be submitted in duplicate, typewritten on one side of the paper, with wide margins and double-spaced throughout. Submission of final edited copy on computer disc will be requested, but two hard copies of the text are acceptable if computer facilities are not available. Format should follow that used in recent issues of Watsonia. Underline where italics are required. Names of periodicals should be given in full, and herbaria abbreviated as in British and Irish herbaria (Kent & Allen 1984). The Latin names and English names of plants should follow the New Flora of the British Isles (Stace 1997). Further details on format can be obtained from the Hon. Receiving Editor or by viewing the website at: http://members.aol.com/bsbipubs/ watsonia.htm. Tables, figure legends & appendices should be typed on separate sheets and attached at the end of the typescript. Figures should be drawn in black ink or be laser-printed and identified in pencil on the back with their number and the author’s name. They should be no more than three times final size, bearing in mind they will normally be reduced to occupy the full width of a page. Scale-bars are essential on plant illustrations and maps. Lettering should be of high-quality and may be done in pencil and left to the printer. Black and white photographs can be accepted if they assist in the understanding of the article. If you are able to submit figures on disc please contact the Receiving Editor to check they are in a Suitable format. Contributors must sign a copyright declaration prior to publication which assigns the copyright of their material to the Botanical Society of the British Isles. Twenty-five offprints are given free to authors of Papers and Notes; further copies may be purchased in multiples of 25 at the current price. The Society takes no responsibility for the views expressed by authors of Papers, Notes, Book Reviews or Obituaries. Submission of manuscripts Papers and Notes: Mr M. N. Sanford, c/o The Museum, High Street, Ipswich, Suffolk, IP] 3QH. Books for Review: Mr D. A. Pearman, The Old Rectory, Frome St Quintin, Dorchester, Dorset DT2 OHF Plant Records: the appropriate vice-county recorder, who should then send them to Dr C. D Preston, Biological Records Centre, Monks Wood, Abbots Ripton, Huntingdon, PEI7 21S. Obituaries: Mrs M. Briggs, 9 Arun Prospect, Pulborough, West Sussex, RH2O IAL. Back issues of Watsonia are handled by Dawson UK Limited, Cannon House, Folkestone, Kent CT19 5EE to whom orders for all issues prior to Volume 22 part 1 should be sent. Recent issues (Vol 22 part 1 onwards) are available from Mr M. Walpole, B.S.B.L, 68 Outwoods Road, Loughborough, Leicestershire, LE11 3LY. Watsonia August 1999 Volume twenty two Part four Contents Kay, Q. O. N., Joun, R. F. & Jones, R. A. Biology, genetic variation and conservation of Luronium natans (L.) Raf. in Britain and Ireland Barratt, D. R., Wacker, K. J., PyweLi, R. F., Mountrorp, J. O. & Sparks, T. H. Variation in the responses of infraspecific variants of wet snes, species to manipulated water levels Preston, C. D., HoLLincswortu, P. M. & Caen R 7 The Ceca a ne of Potamogeton x suecicus K. Richt. (P. # — Pers. * P. eis” L..) im the British Isles ee Wricut, J. A. & Lussy, P. S. The - ana sieneiat status a Manes uniflora (L.) Gray (Pyrolaceae) in Scotland ee Carey, P. D. Changes in the distribution and dines of Himantoglossum hircinum (L.) Sprengel (Orchidaceae) over the last 100 years KircHNER, F. & Buttock, J. M. Taxonomic separation of Ulex minor Roth. aaa U. gallii Planch.: morphometrics and chromosome counts Ric, T. C. G. & Lewis, J. Use of herbarium material for mapping the ‘ectdna a of Erophila (Brassicaceae) taxa sensu Filfilan & Elkington in Britain and Ireland WHEELER, B. A. R. & Hutcuincs, M. J. The history and distribution of Phyteuma spicatum L. (Campanulaceae) in Britain Ricu, T. C. G., Lamprick, C. R. & McNap, C. aurea of Britain’ biodiversity Cyperus juseus L. (Cyperaceae), Brown Galingale Rich, T. C. G. Conservation of Britain’s Bite § Salvia pratensis ‘ (Lamiaceae), Meadow Clary Rumsey, F. J. Rumex x akeroydii — anew feck hybrid McKean, D. R. A new Susana sbi from Scotland, E. aes: A. Cunn. x E. montanum L. ; so) |, ae Notes Cook, P. J. Elytrigia repens (L.) Desv. ex Nevski ee arenosa (Spemner) A. Love (Poaceae) in south-east Yorkshire (v.c. 61) : Rich, T. C. G. The potential for seed dispersal by sea water in Eagiee wie (0. E. Schulz) Stace and C. monensis (L.) W. Greuter & Burdet subsp. monensis Welch, D. & Innes, M. Southward recolonisation by Mertensia maritima (L.) _— on the coast of north-eastern Scotland seh Edgington, M. J. Erica ciliaris L. (Ericaceae) discovered in thee Bieideen Hills. on the Somerset-Devon border (v.c. 3) Holyoak, D. T. Gentianella oe celica ) Borner (Canines iiliecanee in north Devon es Allen, D. E. Rubus campaniensis Winkel ex Beek Baten: in Britain Porley, R. D. Separation of Carex pee L. and C. otrubae on eke ; using transverse leaf sections ; Book REvIEws OBITUARIES Published by the Botanical Society of the British Isles ISSN 0043-1532 Typeset by D. K. & M. N. SANFORD Printed in Great Britain by THE BOOK COMPANY, PO BOX 243, IPSWICH. SUFFOLK ai Nien ewe Wary LOL EIU Il 3 9088 01167 5394 SMITHSONIAN INSTITUTION LIBRARIES cee pate Ug tg oS fo pot i sete po thy a Mere oT! oN aes Pee Seppe tome pein Meine y OR AEA ETD Pee ad <4 Pepe et eee heen ete i Shes et hema OR UN ge LP EN GM ON nite’ Pe Osta Ae oP INS OM, sey wis AT Wether sam erie aa se iealiad nee PC ph AAU ets poem Mai A AMES PTT PE Wee V a ee fine We Lee es em, Puna i . marinara tT Re LEGS NFP efemaitetsiner eas nan See LCA 8) Satta err eT Tad BAYT ORIN RP ey Pte | al ee ae pennant NON a ene ENE TIE ENON eis MAMA RA NAMEN wee content whee rine wank peewee PIMA TENA De EEE oy Oe ‘ nead ph LINEA AARNE EMANY VTo8 ZINE AOR MITE mY he RE aR A SP ‘een eos HUD OPME AT RONEN EIN SA peel IV CANVE Om ONANAM ON PN DAES TUT 7 Aedes reef hen CP at MPG SEN ai nate win eel a ERA Mie UPA seed myeiye, ras Where ere oa ne ad Gree eC ey a) Arai RENAN EAS Ya VEE Len Et Aa a OF git hme ONS CAMEL Nerasmytens Pure i sy rhe eee SOUP Ah at Sy, eyes mating PAE OOM ING 1A By PE WADED GS Phe BPO Ne DA NEAL he Sle Coe Vee IN GI NMRA ME Se iam BAM TLD LEAL SAN UA EME PEE NORE OER BEANE EME SALARY TAR enadewin Gr Wah EMC BD OE! ta tbat Sats a7cest te Ca SAA PARLE co ea IME tot 13 UU WEL ati A rile LL dep Me APSE METAL LO ALD ORME LES WOM T EN Nae EN VERO suas Wace wv b vgtes wove Uh ZN nana ANE ENE NI Nt at UR MAT STL M mA oS HLT WERE NE UA RRIES had rate OME LOE EME NPR mn Haar Sakai oN BANE EY a vente get ares phete GUA PREMISES Fos neta RDS A fae Ae SLE YANDEL LNA TP RAUNT TIA RE NORTE ree Oe REE AA MENA EOF ed Sv fe RUE EAD re ad (oni yf ERT Aa AS aT Ae Np tpn ore eV EN OSE renee sag YAN. feeat@ALe NEN OT eel a LAG RT IL Cra Cab AIAN iL GARD AEE PO TU ae ee a di ee tip mag NAAN sient wyeiyss ridin AD ynpeneit ba ler ne’ nt wana ct cen et a ae el teenie rein hewn Pads fiiawriaee nso cae PE Cn a 4 at ve Mow yer Heng stent of AN ARERR SEP Bs 2 A ere ans LEMON PAWL IS NELE RU TLy) bina: DAE eae SOURS ENED NIM GH Perino AAW eet uate sinfonh aie Cath Me PAA TENE TENE Pye oe un uys ud eZee PLA ene sy ta baaees EVENT when en? Tsadaee tae ut wed a ok Oe Mm RAIA’ 2} SP ANE NEMS Nene ala, Fait oft al aed efo ads ane Antes SEAMEN ce HY KT AH a VRID AME Sc WOM NDA LEEE AIDA PAAR Pagenbiy AES oe aura URE CGN MSU SAMOS RAB a UE outa et eaten ene eCPM E OS alba DMA BANE Be Vag SEAT NNI NT AWD RENE GE feast Fae ier etre ee ae rou icant ie UE rn MEA een ANNA a CEI INAH ae al et a ee ayes agp teadt Matt % len re vacant nike igeaionr it g OR ae ee staal Aap iat NEE ad MAIC SD aan atne Mm A ye SPAS E aed tg tee HM Se AU a aif Meno. Hae 20 8 PRET i” cmicnine ret zirelith bane radia bang eign eT oy AeA ha hee seater dene WsinAnaadiea'yals wet parton h Coteat he LARVAE AI ene rant ve cing Rane thy NWP yh Se aA ap bee OU AE OT MES, UP Ut NE ORS i DUMONT a NT Go eae nea aay! Te on Ve a by triat yy dy asd fens RUA CT te? Re AMAT NDT EN nanan 1B 0 8h ea eles apa rnin wala) 9 re PUnineda oe Un Ponta (wits nn es nC a an ea ee in AeA AT Rad i OME MP MDA TE NUN OEE A aK PPADS POS paca be ek EPMA COON RUSE Lites FE CORY At 81a Abs tha AA bonus diy paras 3 Wires arab iely, ‘ re Fy GEMS A Eten a piel Ay Aer Tee Ep Gineie Cane EOD MENS got ta WOMEN OT Oh epidetee tote Be aspnata lewd Capea hn tat aneh® LED Aan N EAMG OES 748 LES MEA DEE, po he A nhersbnnyma madame aeam nn Ware AN at WE AIAAY Vey Ee On WAP EME Be ahve hl ole tht ce VUE A NERD " UNAS ANA Bes BQ te AD 8S i qo 5 Fal ae nde Rimataulen dete i ancien ait ee NPAT ISG © re ghe (Oe UE ade a dee Sept as ae LAM ity PARAM MEN An NTR ENA DO RTE DEAL rh Sey yee ce Le ea LPP RM OD A VELUN ENE fe ; paakelt yn HED PNT rsa ; H Ce i ee ne ardyneos Fee A NE AN Pee en a te AS SSN oe tenet wh ls pA fet 2 9 Alo ha ae pe EET lf aah ENP PVA ee eS Ly heeet WAT UA EAL MEATS hagees MNBL LAP oH pa je RATA TAN sth We fm petit ae esta BME D AL Mbijels Ra ras cifsFI fhe tne yeb ay Naecibanedligye haatieneiiset Net as a sain eer ATE 0 (Me wh ta SIA Pek veh He ee hl et oe pe a ROAR PE CE Ry LRN OE AL Sete Ah ACO UL Pn PANN 1 at ITY WAR TAS ES PMG my LAGS te Rib eEP A iprisaim lym hy unten det . rary a ae Lh a sii LAV iie es Aiea at one prea NOVEL AE ORIAT NCAT ined ed bry wan baba nt adelbsre aba CPS, COVEN AD PALANAN E pepe ton ki AA ASL SELES LINE GS EY 21 ME TEES TT ea 1 a ora y Ai Hasan A ues ae AYA eg SN iyed yaa One sisrtagnimis Fo eh iP hd ak iol es reas Op A ietth OF si ph dr esee Use uy tag Pt Whee Nip ibaot 0M Myain Dig ah foul Leqie meat AE pate Lets sain bi a's: aaginmgsy di eae . eahian Senne ay Crate NA SEES th aa tram Mop ; 4 a a atinth AN CUT eee over il eda aunties a Wee NDNA ede 8G le nya 4 " if n tf heed teal sdea Whar U8 & ee eda perl cE sb be Sab reds eet ia me eh ihr WN ci ttiasselaea aii oy NA URL ENP CDRA me TBA PALE IS setaaas wt A tt ev AAR En ooattaeth ee he Wag AeA ve Ai ee Ey yea wast dad § COAL EAM ty 08 ld avatar AEN ETN A ROP A BAL Te Aevagi nk bie Metis Nia 9 MLA by Miia toy BLU ARAL Be aad ie viable MAE ENN fen eS ME AY Obs we Ki ey eD raet tibet nei a Bie akond oer WEA LAAT URT AEN OER ACMA NTIS Ag IND Ay TEU UA Sy A Ue We NAL Ard ps pues ttonl MONEE Me Dla BA ENE Cop Hats, SEED SOND AY GE M Dai , 1 ty snenolns ate sa sely tt Hed ah uae APIO PAS YAMA ES Me RIV IN CBO LNT SR ae DED MOND AW SMEG TES IN Ce analy da dln ae, fant We dn» cs at te sea naise veneas Coen ve NPAT aes ay aia ai Verb Wiens bs 1 Pere BURN IY aE A Dl a a Kenran Re Ly aspenvionsienlg ce) EME ED APSE MM NEN ID Thad NEALE NEAL AD RAVE Sah Ayes alp net S08 100 WAN Toei si sina A tapes ih pail aN tae ovata “hs sbebae haa rppoMpse nia peices | Wb LRP NE Nang. Waele 4 ane! F ahenbitlh i} Mite Coes Ad OTA LEO Ae, Ninh a PATS EERE 0D, lobia tind ea dette fa By Une Zit ip AU SP UE NAD Tah Mk eAltew eed Payee Pine ee Bo SVN Pa 708 Rh © bape NUR od woke ord NG bes path nia het ryt Ms Lap SPA 8 ey ae AS ELI EA I TG i orn Bri Han A wads Aan anager wads ve itd ae’ (ED Veale LDN a AEM ( ig Oa hehe riers Leah een ia Ine Leia ay nies VitG eas ween smite erie es Att piaks AOE Pit wren tbat safhine HUAN sheen th APM AAG I AML ga RMA Ah MIE RIO A AIREY HELIS pelt MANA Aibeepinteret rc tt eae i) be ENA, We ehe ide Le a wp ‘ sil i ylen “s NOR SHE ANF srry ROA PR LPARAM i Se WIAABIN AY AS AQMN GS ANS al Wd Ueiayy 38A AAA NDIA OPIN LL 9 PP MMSE ate SWAP a & Keren " Wem ay ag ma em AVN Aang NESS CAN SET AN Hp nataet PE Range a oak nds Resetin at | eons lam ery HMMS E hed oo a MUD Ot Ae Syltaya hae at ten gn nea Wench atime nd " LAM a ie Nesiarin et me? recA al al NN ac Weenie an Pk ONE te ay ot ee MEN IE BPR EME ME LE om Leet: vei s if ae ei Xi esnetd genta eileen bate Be Sinienistired snap sig brtcios aa caab er be engt Ae Abels ie, et HGS pad Ne =i celina een WeieMiAat peibepen ire) BACAR far a ney eure A ait die si aigennwinia gel AHS 85 ED ep RA et g BET HEIN WL iy Cass ae are fe eee ANAL RPL TIDES TSO Rey PARAS HS Bal bel fa al i ak Bsa Alyn Qs gma ahaa HY 9, eh Otel eta tetle Ae ihe ded Mean Sel Mt dryer FG Cyn ALA NEAT NESS W GEA GENO Abe ps BN GN Naiman (fone ENON Me Hib A re NUitib dl. i an eer D ALMA Abe MS rere SY MUCH EL pruery